HA43 : Selecting and Optimizing Portfolios Using mextaheuristic Methods and Comparing Selected Portfolios with Stock Market Expert- Amateur-Selected Portfolios in Tehran Stock Exchange
Thesis > Central Library of Shahrood University > Industrial Engineering & Management > MSc > 2010
Authors:
Arash Talebi [Author], Mohammad Ali Molaei[Supervisor], [Advisor]
Abstarct: Contrary to the growing use of portfolios and in spite of the rich literature on the subject, there are some problems and unanswered questions. Besides, Iran’s stock markets, as developing and growing maekrts, need native researches to answer the problems and unanswered questions. The aim of this work is to be a useful instrument for helping finance practitioners and researchers with the portfolio selection problem. While reviewing Modern Portfolio Theory’s (MPT) literature, this study describes the advances and developments in the field of portfolio selection and optimization, and also investigates optimization methods. Then, heuristic methods are determined efficient methods due to the advantages they hold, and therefore, four of them, which happen to be of the most efficient and newest ones, are selected to resolve the portfolio selection problem with the objective of simultaneous risk minimization/return maximization in Tehran Stock Exchange market and from the stocks of listed 50 top companies. Moreover, to investigate the variable of input data (Annual vs. Monthly) on portfolios performance of efficiency and effectiveness dimensions, two different portfolios are selected via each algorithm and each input data (Annual vs. Monthly). Thereafter, in order to survey the performance of algorithms, experts and amateurs altogether, experts and amateurs are defined by the author in the first step and then their would-be portfolios are gathered using a questionnaire. All portfolios of this research, including eight portfolios of algorithms-input data, forty of the brokers as the representatives of experts and forty-three of individual investors who are present at stock market, defined as amateurs, are applied to the real data of stock market for a six-month period, called test period; in other words, in real market, stocks are bought and hold hypothetically for six months according to the portfolios in an inactive strategy. Finally, the performances are calculated according to risk-adjusted indices, and then, main and subsidiary research hypotheses are examined accordingly and via ANOVA and Scheffe’s Post Hoc tests as statistical procedures. The results indicate that there is not a significant difference between experts and algorithms’ performance. Moreover, both have achieved better returns in contrast to market portfolio during the test period. Algorithms’ convergence speeds are also reasonable. But, amateurs’ mean performance is in a significant difference in contrast to the previous groups, as it turns out, by performing Post Hoc test of Scheffe, experts’ and algorithms’ performance was significantly better than amateurs. Findings also show that the variable of input data, did not have any effects on the portfolios performances. Thus, the conclusions are: 1. As the algorithms were compatible and consistent with the portfolio selection problem, experts, who are invoking enormous human and financial resources to construct portfolios, are advised to employ the algorithms instead. Following this recommendation, same effectiveness is resulted while leading to more efficiency. 2. Amateurs are highly recommended to construct and hold portfolios instead of individual stocks, but as they were shown not to have a good hand in portfolio construction, using the algorithms is a wise action, at least in the beginning, those who are not capable of constructing portfolios, are leaded to buying investment corporations’ stocks to be a part of their portfolios indirectly. 3. As the input data did not cause any difference in performance, using annual input data is advised to the researchers and investors; because it needs fewer calculations and therefore, more efficiency while maintaining the same effectiveness in contrast to its monthly counterpart. 4. As the algorithms only used historical data to construct the portfolios, and they achieved a good performance in contrast to the experts and the market portfolio, the weak form of efficient market hypothesis, which approves digestion and reflection of historical data in stock prices, seems of no effect and is under question in Tehran Stock market.
Keywords:
#Portfolio management and optimization; Modern portfolio theory (MPT); Evolutionary and heuristic optimization; Genetic Algorithm; GA-Nelder-mead hybridized algorithm; Particle Swarm Optimization (PSO); Imperialist competitive Algorithm; Stock market Experts; Stock market Amateurs. Link
Keeping place: Central Library of Shahrood University
Visitor: