PORTFOLIO OPTIMIZATION OF ASABRI AND JIWASRAYA USING FIREFLY ALGORITHM

• Main Author: Andriani, Rosa
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/71787
• Institution: Institut Teknologi Bandung

News at the end of 2019 regarding the Asabri and Jiwasraya has become a mega scandal for state-owned enterprises of Indonesia. This case caused losses to the country's economy up to trillions of rupiah due to non liquid and penny stock investments. Non liquid stocks are stocks that are not actively traded, marked by the fact that there is not always a queue of orders for price fractions at the bid price or offer price. Penny (small cap) stocks are stocks with high volatility due to engineering carried out by market players for a specific purpose. As an investor, you will usually avoid these two types of stocks because they are too risky, do not have good fundamentals and being mark as unusual market activity (UMA). Therefore, this research examines the optimum risks and returns of Asabri and Jiwasraya’s portfolio using the Markowitz Portfolio Model. The problems studied are single-objective and multi-objective problems. In single-objective portfolio optimization problem, we look for the proportion of stocks that will generate the minimum risk and also look for the proportion of stocks that generate a predetermined maximum return. For the single-objective problem, the constraints used are buy-in threshold, namely constraints that limit the minimum value of the proportion of stocks so that the proportion obtained is not too small, cardinality, namely constraints that limit the number of stocks to be included in the portfolio, and roundlot, namely constraints that allow finding the optimum value of investment stocks in units of lots. In the multi-objective portfolio optimization problem, the optimum value of risk and return is sought simultaneously using the sum-weighted method. 21 weights are used to build a pareto front sigma to return, where sigma is the square root of risk. On the pareto front chart, known that a high return accompanied by a large risk as well. Because there are differences composition of stocks at the time of the case released and the latest conditions. Then the two data are used as a comparison. The data at the time of the case release were 14 Asabri stocks and 9 Jiwasraya stocks in April 2019 - December 2019, and 30 IDX30 stocks in January 2014 - December 2019. And the latest data is 20 Jiwasraya stocks in January 2019 – November 2019, 21 Asabri stocks in January 2018 – December 2018, and 30 IDX30 stocks in January 2017 – December 2021. The IDX30 stocks use the multi-objective portfolio optimization problem with cardinality (K=14, K=9, K=20, K=21). This was done as a comparison of the results of optimizing returns and risks for the Asabri and Jiwasraya stock portfolios. This portfolio optimization problem is solved by using the Firefly Algorithm (FA). Firefly Algorithm (FA) is a metaheuristic method developed based on group intelligence intended to mimic the characteristics of fireflies. To evaluate FA performance, three benchmark functions are used, the MINLP problem, the portfolio optimization problem with buy-in threshold and roundlot constraints.