A Portfolio Optimization Model for a Private Equity Investment Company under Data Insufficiency Condition with an Artificial Bee Colony Meta-heuristic Approach

Different investors with different investment levels have a goal in common which is to reach a portfolio of assets which further to meeting the expected rate of return would have the least possible level of risk. In this study we aim to help an investment company to determine an optimized combination of assets containing the stocks of its subsidiary companies as well as other lower risk assets. One of the main challenges in investing in private companies’ stocks, is the lack of data related to their return and risk compared with public companies. In this paper we apply a simulation approach which is able to generate valid random numbers in data insufficiency condition to calculate the return and the risk of the private assets. Furthermore, defining the problem as a bi-objective optimization problem and regarding the fact that portfolio selection is an NP-Hard problem, we use a multi-objective covariance-based artificial bee colony algorithm to solve our problem. The results show that efficient portfolios are the ones have both high risk and low risk assets simultaneously.