• Text Is All You Need: Asset Pricing Without Returns – Job Market Paper  (Link)

How should investors value firms without return histories? In standard valuation practice, discounted cash flow models require a firm-specific cost of equity, yet this cannot be estimated when trading data are missing. Investors therefore rely on public peer-based betas as coarse proxies for systematic risk, introducing valuation errors whenever firm-specific exposures differ from peer norms. Using IPOs as a natural laboratory, I show that textual risk disclosures can substitute for missing return data. I develop Aggregated Cluster Embeddings (ACE), which convert qualitative risk narratives into structured firm-level representations. Disclosure-based betas reduce market beta prediction errors by up to 27 percent relative to peer-based betas. However, investors underweight this information at issuance, and firms whose disclosures imply lower risk than their industry peers are initially undervalued, yielding monthly six-factor alphas of 97 basis points. These excess returns fade as return histories accumulate, consistent with markets gradually learning firm-specific covariances. The results highlight a mechanism of information substitution and investor inattention in pricing firms without trading histories.

• Global Business Networks (with Sebastian Müller), Journal of Financial Economics (2025). 

We generate historical business descriptions with LLMs and construct a time-varying global business network using embedding models. 

• Automated Stock Picking using Random Forests , Journal of Empirical Finance (2023).

I apply a machine learning model to technical indicators to calculate the outperformance probability of stocks.

• Macroeconomic reports and the cross-section of industry returns – Working Paper (Link)

with Sebastian Müller and Garvin Kruthof

• Machine Learning the Impact of Climate Change on Firms Worldwide – Work in Progress

with Gerard Hoberg and Sebastian Müller

• Machine Learning the Performance of Mutual Funds on a Global Scale – Work in Progress

with Manuel Mazidi, Sebastian Müller and Florian Weigert

Five years of teaching experience since 2020. Selected courses include: 

• Investment and Financial Management 

• Digital Finance

• Advanced Seminar in Finance and Accounting 

• John A. Doukas Doctoral Best Paper Award

Text is All You Need: Beta Estimation Using Aggregated Cluster Embeddings, European Financial Management Association, 2025

• Doctoral Award

Friends of TUM Association, For an outstanding dissertation in finance