Optimal hedging af valutarisiko i internationale aktieporteføljer

Christian Siboni Ottosen & Lasse Wagner Olsen

Student thesis: Master thesis


This thesis investigates a new approach to optimal currency hedging for international equity portfolios using a modified mean-variance optimization model (MPMVO) developed by Boudoukh et al. (2019) from AQR Capital Management. The model is designed to enhance the practical applicability of traditional mean-variance optimization (MVO) for hedging currency risk with implementable portfolio characteristics and robust solutions. The model divides the optimal portfolio choice into three sub-components: an initial equity component, a currency hedging portfolio that aims to minimize the equity volatility and a currency alpha portfolio that aims to maximize the risk-adjusted returns through well documented factor investing. The main purpose of this thesis is to provide an overview of the theoretical and practical relevance of the MPMVO model for optimal currency hedging in international equity portfolios. We derive the model and outline necessary theoretical building blocks in the design of the optimal portfolio. The underlying theory will determine our choice of currency components added to the pure equity component. Most empirical studies on optimal currency hedging are limited to the perspective of USD based investors. We provide empirical evidence for optimal currency hedging with the MPMVO model from the perspectives of EUR based investors. We find that the combined MPMVO portfolio outperforms both 100% hedged and unhedged equivalent equity portfolios over the sample period with lower realized volatility and higher excess returns. Finally, we introduce implementation considerations and results on model robustness. We document robust empirical results on the impact of 1) changing model inputs such as rebalancing frequency and leverage, 2) subperiods, 3) transaction costs and 4) short-term shocks exemplified by the current coronavirus crisis.

EducationsMSc in Business Administration and Management Science, (Graduate Programme) Final Thesis
Publication date2020
Number of pages134
SupervisorsSøren Bundgaard Brøgger