Not logged in.
Quick Search - Contribution
Contribution Details
| Type | Master's Thesis |
| Scope | Discipline-based scholarship |
| Title | Four-moment optimization-based portfolio allocation using recent techniques: An empirical approach |
| Organization Unit | |
| Authors |
|
| Supervisors |
|
| Language |
|
| Institution | University of Zurich |
| Faculty | Faculty of Business, Economics and Informatics |
| Number of Pages | 114 |
| Date | 2019 |
| Abstract Text | This research is an addition to the four-moment optimization-based portfolio allocation literature. In an empirical approach, four estimators developed by recent researchers (Martellini and Ziemann 2010, Boudt et al. 2016, Cornilly and Peterson 2019) are used to approximate variance-covariance skewness-coskewness and kurtosis-cokurtosis parameters of the return distribution of three different data sets. The fundamental objective is to assess the impact of the moment-comoment estimation on the expected utility in three different universes and dimensions. In a portfolio optimization allocation using a Taylor series expansion, we found that portfolios constructed with naïve estimators perform well in a low dimension universe and badly in the case of a higher number of assets. In addition, portfolios built up with structured estimators are subject to low performances in period of volatility peak and high correlation such as the Global Financial Crisis. Moreover, the use of shrinkage estimators to optimize the allocation of random variables enables to get a constant good performance in each of the three data set. We also pointed out that the coefficient of risk aversion of a rational investor has no impact on the outcome of the optimization regardless of the type of asset and the estimator employed. Finally, the conclusion advanced by researchers that when assets returns distribution deviate from normality naïve estimators generate the lowest utility could not be clearly verified in the empirical analysis results of this research. |
| Export |
BibTeX
|