Equity Portfolio Construction And Optimization With Applications

The notion of diversification in investment and portfolio optimization was first formulated into a quantitative portfolio theory by Markowitz(1952). This work was then called the Mean-Variance Model of the classical portfolio theory. More than a half century since then has passed, virtually all the follow-on research works in portfolio theory can be considered as deepening and extensions of the classical portfolio theory.This thesis advances a quantitative and dynamic methodology for equity portfolio management, which consists of two innovative processes – stock selection and portfolio optimization. The methodology is motivated by an investment strategy that a general equity portfolio should be constructed first on the stock level and then considered on the portfolio. For stock selection, a set of corporate financial ratios are carefully chosen and used for stock screening. Selected stocks are weighted into a stock portfolio through a multi-objective decision process and minimum semi-absolute deviations.This thesis provides a literature review on the modern portfolio theory, value and growth investing and multi-objective decision making. The research problem of the thesis is then clarified as how to construct and manage an equity portfolio in the dynamic market environment.This thesis discusses investment strategies for stock selection based on corporate value information fusion. A set of corporate financial ratios are selected, based on which an approach is established for stock screening. Evidence from Shanghai stock market on stock screening with this approach is provided.This thesis describes a new model of multi-objective decision-making for determining the weight of stocks in a portfolio, which combines the constrained fuzzy analytic hierarchy process(CFAHP) with(TOPSIS),. This combination overcomes many shortcomings of CFAHP or TOPSIS if used alone.When using the TOPSIS method separately,alternatives’ attribute weights have strong subjectivity. This combination method obtains the attributes’ weights by using the constrained fuzzy AHP to deal with the triangular fuzzy number, and meanwhile overcomes the invalid results by application of ordinary operator into fuzzy number. The combination method obtains the objective weight vector.In the combination method, the TOPSIS method is applied to rank the alternatives and reduces the huge computation in traditional AHP method.This thesis advances a portfolio model based on minimum semi- absolute deviations criterion, in which a multi-factor model is included. This portfolio model has advantages over Markowitz’ s mean-variance model in that only downside volatility is treated as risk。 In the literature, many portfolio models do not consider the impacts of the external economic environment. This thesis has found out a set of macro-economic indicators as influence factors that are included in a multi-factor model as a part of the portfolio optimization. This portfolio model has incorporated the transaction costs into the optimization process, and it can deal with multi-stage portfolio optimization. The methodology of portfolio management developed in this thesis goes beyond portfolio optimization,reaching trade management. To avoid large losses dues to dramatic price fall, a stop-loss mechanism is implemented to manage open trades.This thesis documents the implementation steps and the results of empirical test of the proposed portfolio models and methodology on the historical data. First of all, choose suitable stocks from the set of corporate financial ratios; secondly use the combination of the multi-objective decision-making method and minimum semi-absolute deviations portfolio to do a research, the study results show that the method is effective and practical.It also provides a good theoretical support for investment decision. At the end of the article, the full text is summarized, and put forward the direction of the further research and thinking.In summary, this thesis has proposed and developed a comprehensive methodology for quantitative and equity portfolio management. The contributions of this work include four points: 1) stock screening, a set of corporate financial ratios are carefully chosen and used for stock screening; 2) a multi-objective decision-making procedure for stock portfolio weighting; 3) a portfolio model based on minimum semi-absolute deviations criterion with a multi-factor model for sector index portfolio optimization; 4) a stoploss mechanism introduced to manage open trades as an extension of classical portfolio theory.