Available Topics

Our chair currently offers the following topics for Master’s theses:

• The Transmission of Monetary Policy Shocks (building on Jarocinski and Karadi, 2020, or Miranda-Agrippino and Ricco, 2021)
  [Monetary policy shocks are increasingly identified through the use of high-frequency instruments around central bank announcements. High-frequency movements in asset prices in a small event windows (~30mins) around central bank announcements convey information which arguably stem exclusively from the change in the monetary policy stance. This thesis makes use of one of these high-frequency instruments in a proxy-VAR (or proxy-LP) framework to estimate the causal effects of monetary policy on macroeconomic quantities. It is also possible to extend and allow for non-linearities in the estimation. The topic is well-suited for students interested in empirical macroeconomics, causal identification in macroeconomics, and time series analysis.]
• The Macroeconomic Effects of Fragmentation (building on the index by Fernández-Villaverde et al., 2024)
  [After a period of globalization, the world economy is becoming more fragmented since the late 2010s and early 2020s. This topic is thus well suited in the newly emerging field of geoeconomics, which links economic with geopolitics. Which implications has a more fragmented world economy on trade and economic activity? This thesis will use and explore the causal effects of fragmentation on macroeconomic quantities. The exact research question can explore different dimensions and settings in international macroeconomics (e.g., geographical coverage: world economy, single- or multi-country settings; target variables: exchange rates, economic activity, etc.). The topic is well-suited for students interested in empirical macroeconomics, causal identification in macroeconomics, and time series analysis.]
• The Macroeconomic Effects of Global Supply Chain Disruptions (building on Bai et al., 2024, or Finck and Tillmann, 2023)
  [The world economy is organized around an intricate global supply chain. Any sudden and large shocks to this global supply chain (e.g., the COVID-19 pandemic, or the Red Sea Crisis) might have ramifications for the world economy. This topic is thus well suited in the newly emerging field of geoeconomics, which links economic with geopolitics. This thesis will use and explore the causal effects of global supply chain disruptions on macroeconomic quantities. The exact research question can explore different dimensions and settings in international macroeconomics (e.g., geographical coverage: world economy, single- or multi-country settings; target variables: exchange rates, economic activity, etc.). The topic is well-suited for students interested in empirical macroeconomics, causal identification in macroeconomics, and time series analysis.]
• The Use of Bayesian Shrinkage Priors (building on Giannone et al., 2022)
  [The use of high-dimensional data sets (‘big data’) is increasingly proliferating in economics. Regularization via Bayesian shrinkage priors are thus a useful extension of the econometricians’ toolkit. It enables the researcher to differentiate signal from noise and ‘let the data speak’. The thesis can explore a high-dimensional data set using a well-established shrinkage prior (e.g., Stochastic Search Variable Selection prior, Normal-Gamma prior, Horseshoe prior, etc.) for prediction or causal analysis. The topic is well suited for students interested in econometrics and Bayesian analysis.]
• Priors for Bayesian Instrumental Variables Regression (building on Lopes and Polson, 2014)
  [Instrumental variables (IV) regression is used to estimate causal relationships when controlled experiments are not feasible. It allows to measure the causal effect when an explanatory variable of interest is correlated with the error term (endogenous) by using a third variable (a valid instrument) to induce changes in the explanatory variable which are uncorrelated with the error term. IV regression can also be estimated with Bayesian methods. The thesis explores possible prior distributions for an IV regression and applies it to a suitable setting. The topic is well-suited for students interested in econometrics and Bayesian analysis.]
• Building a Business Cycle Tracker Based on Machine Learning Techniques (building on Woloszko, 2024)
  [Understanding the current state of an economy is crucial for policymakers, businesses, and researchers, but traditional economic indicators like GDP are published at low frequency and often with significant delays. This thesis will explore how real-time, high-frequency data—such as Google search trends—can be used to estimate economic activity more accurately and promptly. The thesis will address key challenges, such as the short historical record of alternative data and their complex, non-linear relationship with GDP. Combining approaches from the mixed data sampling (MIDAS) literature and machine learning techniques, the thesis will develop a new high-frequency index of economic activity. The topic is well-suited for students interested in empirical macroeconomics, forecasting, and economic policy analysis.]
• Development of a Data-Driven Sales Projection Model in the Industrial Goods Sector (coopration with Siemens)
  [A data-driven forecasting system is to be developed for the Smart Infrastructure division at Siemens to support the sales department in projecting demand for industrial goods for individual customers. For this purpose, historical order quantities for individual products from a project database (possibly partially automated with the help of text mining tools) are to be aggregated at customer level and updated using simple time series models. In cooperation with the department at Siemens, the aim is to work out how this can serve as a supporting basis for a reporting system with which sales projections can be transmitted from the sales side to the manufacturing plant. The topic is suitable for students who enjoy analyzing data and have the necessary software skills (Python or R), who are interested in the industrial goods sector and who would like to gain an insight into practical work at Siemens. Knowledge of German strongly preferred.]
• A Real-Time Assessment of the GDPplus Methodology (building on Aruoba et al., 2016)
  [GDP can be measured from an income perspective and from an expenditure perspective. The thesis will explore how well the model that uses both measures in a state-space model performed for the US in real-time during the pandemic and in recent years in general. Potentially, the thesis could also apply the model to other countries. The topic is well-suited for students interested in econometrics and empirical macroeconomics.]

Please contact Prof. Dr. Jonas Dovern to inquire whether there are other proposals for topics or to discuss own suggestions for thesis topics.

Potential topics should be related to the content of the Master courses „Multivariate Time Series Analysis” or “Bayesian Econometrics”.