Valentin Braun

Dynamic Copulas for Finance Valentin Braun

The interactions of financial securities are crucial to determine possible portfolio losses. Although this fact is well understood, two questions remain: What causes changes in the dependence structure of financial assets? How can fluctuating dependencies be measured? The most common approach to identify the amplitude of financial assets' interactions are linear correlation coefficients. However, they fail to comprise shifts in the dependence structure. Alternatively, Copulas are a more flexible dependence measurement. This book focuses on the development of Dynamic Copula frameworks by implementing stochastic parameters into Archimedian and Elliptical Copula functions. In contrast to static correlation measures, the Dynamic Copulas are able to replicate unstable financial market interactions. Various Dynamic Copulas are applied to global stock, bond, commodity and exchange rate data to calculate the correlation time paths, which explain financial market reactions to economic shocks. Furthermore, the interactions of dependencies, volatility and returns are analyzed, to determine the efficiency of portfolio diversification in regards to wealth protection. Portfolio risks are estimated through Dynamic Copulas to demonstrate their abilities to replicate financial market interactions accurately. Additionally, this analysis reveals the impact of changing dependence intensities on the magnitude of possible portfolio losses. Finally, the Dynamic Copulas are utilized to allocate higher moment optimal portfolios. This examination emphasizes the effect of inaccurate correlation estimates on the portfolio choice.