Comparison of Performance of Optimal Portfolio based on Various Risk Estimations

Most investors are willing to invest assets which generate higher return at lower risk. For the past decades, many methodologies introduced to measure and forecast the performance of portfolio and to manage risk. Many researcher have suggested various methodologies of performance measures in the investment decision making as an alternative to replace the Markowitz Mean-Variance Model or the Sharpe ratio. Biglova, Ortobelle, Rachev and Stoyanov(2004) compared 11 approaches to risk estimation for the optimal allocations among the risk free assets and nine risky assets in the German market. They reported that the Sharpe ratio criterion under performs Minimax, MAD, and VaR criteria. In this paper, we try to find which portfolio criterion deduces best decision for 3 countries such as Korea, Japan and UK stock markets. Industrial indices rates of return are used as underlying asset s return in the sample period. We found that the performance of optimal portfolio derived from classic models such as Sharpe ratio and Markowitz are inferior to alternative models for each country s stock market in which the return on assets are not normally distributed in common.