Μέθοδοι πρόβλεψης μετοχικών αποδόσεων σε διεθνές επίπεδο και κατάστρωση βέλτιστων χαρτοφυλακίων

View/ Open
Keywords
UCITS ETFsAbstract
The purpose of this thesis is to investigate forecasting methods for international equity returns and to examine whether such forecasts can be used for the construction and evaluation of optimal portfolios. The empirical analysis focuses on ten equity UCITS ETFs traded in European markets, covering a broad range of international investment exposures, including developed markets, emerging markets, the United States, Europe, technology, and factor-based as well as ESG-oriented strategies.
The study is positioned within the broader field of empirical finance, where research has gradually evolved from the traditional single-factor CAPM toward multifactor asset pricing models, such as the Fama-French and Carhart frameworks, as well as toward more recent forecasting approaches that incorporate a wide set of characteristics and more flexible predictive methods. The use of UCITS ETFs is particularly appropriate for the purposes of this research, as the UCITS regulatory framework and the related ESMA guidelines provide a common institutional setting in terms of transparency, disclosure requirements, and investor protection.
The methodology of the study is based on monthly data for the period 2020-2025, on the construction of return, risk, and investment exposure variables, on the estimation of linear and nonlinear forecasting models, and on the assessment of the economic value of forecasts through their application to portfolio construction strategies. Portfolio performance is evaluated using risk-adjusted measures such as the Sharpe ratio, the Sortino ratio, and the Information ratio.
The study aims to contribute both to the relevant literature and to the practice of portfolio management by examining whether return forecasts can lead to more efficient capital allocation and to an improved risk-return trade-off relative to simple passive benchmark strategies.


