Does Naively Selected Perform Portfolio Efficiently? Empirical Evidence from Indonesia Capital Market

Abstract

The aims of this study are to search and examine the optimum number of assets that perform portfolio efficiently. There are 210 nonfinancial stocks listed in Indonesia Stock Exchange included in the samples during period of analysis 2000 to 2008. Monthly returns data (daily cumulative abnormal returns for each month) are used to eliminate bias of non-synchronous trading in data analysis. This study performs 500 naively selected portfolios using replacement random sampling method for each combination of stocks of portfolios from 2 to 20 stocks. Treynor Index and Jensen Alpha are used to measure the portfolio performance. Research results show that portfolio with 18 and higher stocks provides practical consistent value of performance and lower standard deviation. However the results of statistical tests comparing performance of portfolios between and within portfolio j and j-(j-k) do not confirm the hypothesis that the higher number of stocks leads better diversified portfolio. This study finds that decreasing value of portfolio performance slower than decreasing level of deviation standard. Parallel to previous studies (Elton ang Gruber, 1977; Statman 1987; Bennet and Sias, 2006; DeMiguel et al. 2009), this study suggests investors to expand the number of stock beyond 20 stocks to attain better diversified portfolio.