Decision Support System for Portfolio Selection

The developed decision support system allows in Excel environment to describe the set of efficient portfolios and analyze alternatives of investment strategy applicable to individual types of pension funds (conservative, balanced, growth) It consists of the following stages: definition of fund's investment style parameters, optimization of investment strategy, presentation of investment strategy alternatives and their description and analysis of sensitivity of investment strategy to changes in expectations. Within the definition of investment style parameters the following is specified: a) limits for individual asset classes, namely money market instruments, bonds and equity, b) limits also for individual assets, contemplated within asset classes, c) currency (local, USD, EUR, SKK) of individual asset classes, to which yields of individual assets are recomputed, d) expected return of individual assets (historical yields for the whole period, historical yields for the specific sub-period, expert values or long run equilibrium values based on Black-Litterman approach) Optimization of investment strategy consists in effective approximation of efficient portfolios frontier in the space of expected return and risk (measured by standard deviation, lower semi - standard deviation, lower semi-absolute deviation, below target risk or conditional value at risk) and identification of four alternatives of investment strategy, portfolios, having the following attributes: a) portfolio with global minimum risk, b) portfolio with maximum expected return, c) portfolio with the same value of profitability index and safety index, d) portfolio with specified relation of profitability index and safety index. The computed alternatives of investment strategy are presented graphically and described through a system of attributes, specifying their yield and risk-related characteristics, both from general point of view and from the point of view of individual assets selected for the portfolio.