Portfolio optimization techniques

Abstract

Portfolio Management has evolved for the purpose of providing a good match between what different types of investors want and what portfolio managers can offer. Quantitative asset allocation models have played an important part in the global portfolio management. At the higher level in the multi-level approach for investing, investors makes decisions regarding taking positions in various types of investment vehicles. The investment decisions are taken through the analytic consulting of the portfolio manager. At the lower level, investment firms provide a package of assets/securities which can be used by the investors as investment vehicles. Thus the portfolio manager needs to have certain techniques to construct optimal portfolios for different kinds of investors. We have looked at three major techniques for constructing optimal portfolios.1. Markowitz Mean-Variance Approach This is a classical portfolio optimization technique. The use of Black-Litterman approach is also incorporated.2. Conditional Value-at-Risk This approach has been traditionally used by most of the investment firms to minimize the downside risk when the markets are falling.3. Value Investing This has proven to be successful, especially by Warren Buffett, as anon-statistical approach for finding the stocks with higher growth potential. The methodology of each technique is discussed and demonstrated using the data of the stocks from BSE 500 for a period of 66 months (Jan 03 to Jun 08). The pros-cons of each of the approaches and their implications for a portfolio manager are discussed.