OPTIMIZING PORTFOLIO RETURN WITH NAÏVE DIVERSIFICATION-BASED MODELLING

Abstract

This study aims at applying naïve diversification-based modeling in formation of optimal portfolios and to test the superiority of these portfolios against its sectoral indexes. The population of this study are all companies listed on the Indonesia Stock Exchange which are grouped into 10 sectors, namely: Agriculture; Basic Industry; Consumer; Finance; Infrastructure; Manufacture; Mining; Miscelanous Industry; Property; and Trade. The sample of this company is Top 10 Constituents in each company sector listed in the fact sheet per sector, published by the Indonesia Stock Exchange. The analytical tools used were paired sample statistics, paired sample correlations and significance tests. The results shows that portfolio formed with naïve diversification modeling shows its superiority compared to its sectoral portfolio. The correlation test shows moderate significance relationship between returns and standard deviation of sectoral portfolios with naïve diversification-based portfolios, while beta shows no meaningful relationship between sectoral portfolios and portfolios with naïve diversification modeling. Discrimination tests show the significance of returns and standard deviations between sectoral and naïve diversification modeling-based portfolios. While in line with the correlation test, there is no significant difference between the beta of the two portfolios, so it appears that the volatility of the two portfolios cannot be separated from overall market movement. For bearish market conditions, the level of portfolio loss using naïve diversification modeling is lower than sector-based portfolios in the Indonesia Stock Exchange.