A Python Module for Selecting the Number of Assets in Optimal Portfolios via Two Alternative Techniques

The current work presents a software component produced in Python by the authors that constructs the optimal portfolio using two alternative approaches. The first standard approach relies on obtaining the optimal weights via the minimization of the portfolio's variance as a measure of risk. The second approach as alternative method mixes return and risk explicitly in the objective function and gives the needed portion of investment for each asset that results in the highest possible return per unit of risk. The current technique is coherent with the rational expectation that investors combine return and risk once decisions on investment are made. Our software package advances another benefit that is commonly overlooked in practice, which is determining the collection of assets for including in the portfolio endogenously. This software creates all potential combinations and accordingly the investor is able to discover analytically the portfolio that is the best one amongst possible ones. The module offers a graphical user interface (GUI), which makes it consumer friendly. An application is also supplied for demonstrating the way the software operates using real data of financial assets.