Multi objective portfolio optimization for a private equity investment company under data insufficiency condition

Selecting an appropriate portfolio making an optimal trade of between the return of assets and the associated risk of them has been always a fundamental challenge for different investors with different types of assets. The problem becomes more complex for an investor investing in private companies of which she doesn’t have enough data to evaluate its return and risk. Furthermore, this type of investment involves selecting more high risk assets which may not meet the risk attitude of the investor. In this study, a bi-objective portfolio optimization model has been developed to determine the best sets of portfolios for a private investing company. Due to the lack of data on private assets, a simulation based approach has been used to estimate the return of different assets as well as their correlations. A Covariance-Based Artificial Bee Colony is applied to solve the model. The results show that optimal portfolios consist both high-risk and low-risk assets.