Three essays on asset management and capital allocation

Kurzfassung/Abstract

This dissertation explores key challenges and innovations in contemporary asset management through three self-contained empirical essays. Each paper examines a distinct but interrelated topic in portfolio construction, contributing to a more resilient and practically viable approach to investing.

Paper 1: The Performance of Risk-Based Asset Allocation in Downward Markets – An Empirical Examination

The first paper investigates the effectiveness of risk-based portfolio strategies – such as minimum variance, equal risk contribution, and risk parity – during periods of market stress. Motivated by the shortcomings of return-optimized models in crisis periods, the paper conducts an extensive backtest using a multi-asset dataset across several downturns. It finds that risk-based strategies consistently offer superior downside protection and more stable performance compared to traditional approaches. These results underscore the robustness of risk-focused allocations when facing uncertain or volatile markets.

Paper 2: A Performance “Horse Race”: Does Anything Beat the 1/N Portfolio?

The second paper revisits the enduring puzzle of the 1/N (equal-weighted) portfolio's performance. Despite its simplicity, previous research has shown it often rivals or outperforms optimized strategies. Using an expanded dataset and robust methodology,

this paper compares a wide range of portfolio construction techniques – including mean-variance optimization and shrinkage methods – against the 1/N benchmark. The findings reaffirm the strong performance of the naïve strategy, particularly when estimation error and real-world frictions are considered. While some optimized models perform better in specific contexts, none dominate consistently.

Paper 3: Index Tracking in Crisis Periods – An Empirical Investigation of the German DAX Index

The third paper shifts focus to passive investment strategies, specifically the accuracy and stability of index tracking during market crises. Using the German DAX as a case study, the paper compares multiple tracking approaches – such as constrained regression and relative optimization – under both normal and crisis conditions. Results reveal that tracking performance deteriorates notably in turbulent markets, with simpler, constraint-based methods offering more consistent tracking accuracy. This highlights the limitations of passive strategies under stress and points to the need for more adaptive frameworks.