Vliv pandemie Covid-19 a válečného konfliktu na Ukrajině na výkonnosti akciových portfolií

Abstract

This thesis focuses on evaluating the performance of stock portfolios composed of 25 selected stock titles from the DJIA (Dow Jones Industrial Average) index in the in sample and out of sample periods. The out of sample period was substantially influenced by the global Covid-19 pandemic and the war conflict in Ukraine. Portfolio performance was analyzed through various portfolio construction methods, including randomly generated portfolios, Markowitz's mean-variance model, models with risk metrics (Conditional Value at Risk, Maximum Drawdown, Mean Absolute Deviation, Worst Case Ratio, and Ulcer Index), and linear regressions of the Sharpe Ratio, Sortino Ratio, standard deviations, and average returns between the in sample (independent variable) and out of sample (dependent variable) periods. The theoretical framework includes the characteristics of financial markets, regulations, current trends, and challenges, with a focus on capital markets in the USA. The practical part deals with portfolio optimization, calculating performance measures using the Python 3.10 programming language, and comparing in sample and out of sample performances, considering the impacts of Covid-19 and the war. The analysis identified the most relevant models for the given performance indicators.