Modeling and Assessment of Financial Investments by Portfolio Optimization on Stock Exchange

摘要

The portfolio optimization is a complex task, addressing decision making of financial investments. A requirement for the portfolio management is to deal with a lot of securities, traded on the stock exchange, frequently containing hundreds of stocks and securities. The modern portfolio theory insists to be defined a portfolio optimization problem, which parameters have to cope historical trends of the assets' returns, subjective views of experts and additional noisy data, generated from practical and political decisions. The paper addresses the actual model for portfolio optimization based on Black-Litterman approach for integration of historical data and expert views about posterior, future levels of assets' returns. A special formalism is derived, adding additional expert assessment based on the historical trend of the asset returns, which modify the classical Black-Litterman portfolio model. Risk and Return parameters of the portfolios are evaluated in mathware environments, applying classical methods of the mean variance optimization and the new modified model on Black Litterman optimization. By math-ware tools it is demonstrated the ability to evaluate and to assess the portfolio performance. Real data from the world stock exchanges are taken for numerically illustration of the implementation of a modified Black Litterman portfolio model. This research can serve business applications and the topics of portfolio modeling, optimization, decision making, data driven analysis.