Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 Investment Portfolio Management: A Review from 2009 to 2014 Amitava Ghosh1 and Ambuj Mahanti2 This paper reviews the literature on portfolio management from 2009 to 2014 with a focus on the algorithms used, contribution by various nationalities, performance attributes used, etc. and brings out the current state of research in this field. The paper proposes a methodology to conduct a systematic review of the literature and considers seventy-three peer-reviewed journal articles for this. Sharpe Ratio is the most common performance measure. Authors have mostly performed mathematical modeling of the portfolio 1 selection problem and genetic algorithm and fuzzy logic have been used to address the problem. A combination of empirical, theoretical and modeling have been used by most the authors. This review sheds light on how the practical constraints like transaction costs, cardinality constraints, etc. have been modeled in the literature. The review is done on the basis of some peer review journals only. Conference papers and book chapters also have not been considered. Field of Research: Finance Keywords: Review, Portfolio Management, Algorithms, Performance Attributes, Investment 1. Introduction Portfolio selection was originally proposed by Markowitz (1952, 1959). Markowitz formulated the portfolio selection problem as a tradeoff between mean and variance of a portfolio of assets. Keeping constant variance and maximizing expected return or keeping return as constant and minimizing variance led to the efficient frontier of the portfolio from where the investor could choose his portfolio mix depending upon his risk attitude. This has led to the Modern Portfolio Theory. An important implication is that an asset should not be chosen only on the basis of its individual return and risk characteristics but when it is selected along with other assets, its comovement with respect to other assets can lead to same return but can reduce the risk of the entire portfolio. The mean-variance (MV) model has been modified in the literature in many ways. One such work is the market model or the single index model which ignores the covariance between returns of the securities (Sharpe, 1964; Lintner, 1965). The single index model considers the returns of securities to be dependent only on a market index and the co-variance between each pair of assets are not needed. This further led to the development of the Capital Asset Pricing Model (CAPM). Markowitz's single period MV model was later extended to multiple periods (Hakansson, 1971; Elton & Gruber, 1974; Mossin, 1968; Li & Ng, 2000) . Several studies have been done on the Modern Portfolio Theory and the associated risk measures. In addition to variance; Mean Absolute Deviation (Konno et.al., 1993; Zenios & Kang,1993), Gini mean difference (Bey & Howe, 1984; Shalit & Yitzhaki, 1984), entropy (Philippatos & Wilson, 1972), lower partial moments (Price et.al.,1982), Value-at-Risk (VaR) (Gaivoronski & Pflug, 2005), Conditional VaR (CVaR) (Rockafeller & Uryasev, 2002) , Conditional Drawdown-at-Risk (CDaR) (Chekhlov et.al., 2000) have been considered as risk measures. Several papers have considered 1 2 Amitava Ghosh, Doctoral Student, Indian Institute of Management Calcutta, India Ambuj Mahanti, Professor, Indian Institute of Management Calcutta, India 1 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 a combination of two or more risk measures (Alexander et.al., 2006; Bryne & Lee, 2004). A central problem in portfolio management is evaluating the performance of risky portfolios. Several performance attributes have been proposed. (Sharpe, 1966) measured the systematic risk and non-systematic risk of the mutual funds and proposed a ratio to measure the excess return of the portfolio. (Treynor, 1965) measured the performance of a portfolio as the ratio of the excess returns to the systematic risk of the portfolio over the evaluation period rather than considering the total market risk. (Jensen, 1968) proposed the Jensen index or alpha as an absolute measure of the performance of the portfolio manager. If the manager can forecast the future returns better and can achieve better diversification to insure against the risk, then his portfolio will earn more than the normal risk premium at that level of risk. Several other measures have also been used in performance measurement. In order to solve the MV problem, several attempts have been made to solve the single objective and multi objective portfolio selection problem which include linear goal programming models (Sharpe, 1967), genetic algorithms (Xia et.al., 2000; Loraschi et.al., 1995) and support vector machines (Fan & Palaniswami, 2001), fuzzy modeling (Tanaka et.al., 2000), particle swarm optimization (Kendall & Su, 2005), stochastic programming (Samuelson, 1969), etc. All these attempts can be classified into three extensions: (i) developing criteria and models to capture the investor's preferences, (ii) capturing practical constraints of the market into the portfolio planning models, (iii) using features of several disciplines to solve the practical portfolio selection problem. The objective of this paper is three-fold. The first objective is to present the current state of research in portfolio management from 2009 to 2014. The second objective is to develop a framework to perform the literature review of the various papers to find the trend of current research in the portfolio selection field. The third objective is to identify the potential areas of concern based on the review of the papers. This article is organized as follows: The second section discusses the literature reviews in portfolio selection area. The third section deals with the research methodology. The fourth section provides a discussion on the results and the fifth section provides the conclusion and presents the future scope of research. 2. Literature Review (Webster & Watson, 2002) consider literature review as more than just an analysis of collection of summaries of research papers. A literature review can be defined as "the use of ideas in the literature to justify the particular approach to the topic, the selection of methods, and demonstration that this research contributes something new" (Hart, 1998). An effective literature review should create a firm foundation of advancements in the field (Webster & Watson, 2002) and how one piece of research builds another (Shaw, 1995). Very few reviews on various methodologies for solving the portfolio selection problem are available in the literature. But we could not find any single paper which reviews the current trend of research in the portfolio domain considering all the algorithms. This missing link motivated us to carry this study. Table-1 shows a sample of review papers retrieved through cursory search of Google Scholar. 2 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 Sl. No. 1. Author(s) Table 1: Review papers in portfolio management Year Remarks Metaxiotis & Liagkouras 2012 2. Huang 2009 3. Inuiguchi & Ramik 2000 4. Elton & Gruber 1997 Reviewed the current state of research in portfolio selection problems with respect to multi objective evolutionary algorithms (MOEA). Reviewed several definitions of risk in for fuzzy portfolio investment and discussed several credibilistic portfolio selection models along with mean risk model, mean variance model, mean semi-variance model, credibility maximization model, entropy optimization model, game models, etc. Presented a comparison of stochastic programming and fuzzy programming models through their application in portfolio selection problems. Discussed issued related to modern portfolio theory and outlined some important topics for future research. 3. Modern Portfolio Theory: Review of the current literature from 2009-2014 (Chiam et. al., 2009) use the compensatory property of evolutionary algorithm (EA) and particle swarm optimization (PSO) to show its application in computational finance areas which include real world problems like portfolio optimization. It uses PSO as a local optimizer for fine tuning in evolutionary search which improved the convergence ability of the evolutionary search. (Huang and Jane, 2009) integrate the moving average autoregressive exogenous (ARX) prediction model along with grey systems theory and rough set (RS) theory to create an automated stock market forecasting method and portfolio selection technique. (Huang, 2009) integrates fuzzy C-means (FCM) classification theory, variable precision rough set (VPRS), average autoregressive exogenous (ARX) prediction model and grey systems theory for stock market forecasting and portfolio selection. (Lin and Ko, 2009) introduce the genetic algorithm (GA) based portfolio valueat-risk (PVaR) forecasting mechanism using the extreme value theory (EVT). The experiments on the seventy eight companied listed and traded on the Taiwan Stock Exchange show the stability and robustness of this algorithm with success rates higher than the exponential weighted moving average (EWMA) and historical simulation (HS) methods both with 95% and 99% confidence levels. (Soleimani et. al., 2009) present a genetic algorithm (GA) based approach for the Markowitz MV model taking minimum transaction lots, cardinality constraints and market sector capitalization as constraints into account. (Xidonas et. al., 2009) develop an expert system (ES) methodology for equity selection by analyzing the various financial ratios and understanding the strengths and weakness of firms. A significantly large number of firms across multiple business sectors can be evaluated parallely. (Chang et. al., 2009) introduce a genetic algorithm (GA) to solve the portfolio optimization model with different risk measures based upon the MV model by Markowitz: semi-variance, mean absolute deviation and variance with skewness. (Chen & Huang, 2009) cluster equity mutual funds on the basis of four evaluation techniques and groups them on the basis of their performance. The authors use a two stage method of cluster analysis. It uses Ward's method to compute the distance between clusters in the first stage. It then applies the non-hierarchical method, k-means to minimize the variance of objects in one cluster and to maximize variance amongst other clusters. The rate of return and variance are represented as fuzzy numbers to reflect the uncertainty at the evaluation stage. (Chen & Lin, 2009) developed a partitioned portfolio insurance (PPI) strategy by partitioning the portfolio into several similar subportfolios and then insuring the sub-portfolios individually. A new relation-based genetic algorithm is designed to solve this portfolio partitioning problem. (Chen et. al., 2009) propose a portfolio 3 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 optimization algorithm using genetic network programming with control nodes. This method makes stock trading strategies considering the information of technical indices and candlestick charts for efficient trading decision making. (Chen et. al., 2009) distribute the wealth to a set of investment strategies or security rule pairs. The investment strategy problem is formulated as a combination optimization problem and a new combination genetic algorithm is proposed to solve it. (Zhang et. al., 2009) consider a possibilitistic mean-variance utility function and discusses the portfolio selection problem for bounded assets. It proposes a parametric quadratic programming model for portfolio selection for 'n' assets and then presents a sequential minimal optimization (SMO) algorithm to obtain the optimal portfolio. (Branke et. al., 2009) propose an envelope-based MOEA which is combined with an embedded critical line algorithm to solve the portfolio selection problems with non-convex constraints. The EA takes care of the non-convex constraints. The problem is divided into convex constraints which can be solved by the critical line algorithm. The overall solution is a combination of the solutions to the individual sub-problems. (Zakamouline & Koekebakker, 2009) consider performance measure as not just a reward to risk ratio; but as a score attached to each financial asset, which is related to the level of expected utility provided by the asset. The paper presents two methods of practical estimation of Generalized Sharpe Ratio (GSR) and shows how GSR can be used for portfolio performance evaluation by removing the shortcomings of the standard SR. (Diaz et. al., 2009) develop a portfolio selection model to conduct out-of-sample contingent and pure immunization strategies that allow for different degrees of risk and expectations about interest rate movements. (Tiryaki & Ahlatcioglu, 2009) combine fuzzy analytic hierarchy process (AHP) with the portfolio selection problem. The basic approach in a AHP is to set up a model that incorporates the firm's risk behavior (high, medium or low risk), the risk class of the investor (high, medium or low), the investor's objectives and extrinsic and intrinsic factors. (Chang et. al., 2010) evaluate the performance of mutual funds under the broad framework of multi-attribute decision analysis (MADA) where all criteria is considered to provide a final ranking to the mutual funds. (Estahinapour & Aghamiri, 2010) focus mainly on the prediction of future stock price in their study. This paper uses a neuro-fuzzy inference system adopted on a Takagi-Sugeno-Kang (TSK) type fuzzy rule based system for stock price prediction. (Fasanghari & Montazer, 2010) describe a new method for the design of a fuzzy expert system for portfolio recommendation at the Tehran Stock Exchange. It ranks the stock on the basis of fundamental analysis ratios and qualitative criteria from the exchange. The qualitative factors used to evaluate stocks in the paper are arrived at by distributing a questionnaire among experts. (Nanda et. al., 2010) integrate clustering techniques like k-means, self organizing maps and Fuzzy C-means into portfolio management and builds a hybrid system for generating efficient portfolios. (Rodder et. al., 2010) describe a model that does not calculate the classical risk measures like variance and standard deviation. It simply takes the return distributions of the securities as input and suggests an unbiased portfolio by a rule-based inference mechanism under maximum entropy and minimum relative entropy. (Wang & Huang, 2010) construct a responsive mutual fund performance evaluation system using fast adaptive neural network classifier (FANNC) which combines features of back propagation network (BPN), adaptive resonance theory (ART) and field theory. (Alexander & Baptista, 2010) characterize the alpha-tracking error variance (TEV) which consists of portfolios that minimize TEV for various levels of alpha. The paper shows that the portfolios on the alpha-TEV frontier exhibit three fund separation with funds being the two portfolios on the MV frontier and the benchmark. (Anangnostopoulos & Mamanis, 2010) formulate the portfolio optimization problem as a triobjective optimization problem to find tradeoffs between risk, return and number of securities in the portfolio. Restrictions are imposed on the proportion of the investments in assets in order to avoid small holdings or investments in assets with common characteristics. This mixed-integer multi-objective optimization is solved using three evolutionary multi-objective techniques. (Chen et. al., 2010) provide a decision making model for dynamic portfolio optimization and uses the 4 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 time adapting genetic network programming (TA-GNP) which considers the time related fluctuation of stock prices. The paper uses technical indices and candlestick chart as judgment functions for the TA-GNP. (Aluja et. al., 2011) present a framework for portfolio management based on the use of some fuzzy topological axioms and grouping. It uses a fuzzy algorithm that reduces the number of elements of the power sets of related sets by connecting them to the sets that form the topology. (Golmakani & Fazel, 2011) propose a particle swarm optimization (PSO) approach called CBIPSO to solve the extended Markowitz portfolio selection problem. The algorithm is a combination of binary PSO and the improved PSO. The search space in binary PSO is considered as all possible combinations of a two state (0/1) bit string whereas in the improved PSO, a mutation operator is used to search in isolated regions to prevent particles from being stuck in local optimums. The algorithm uses four sets of constraints: bounds on holdings, cardinality, minimum transaction lots and sector capitalization constraints. (Liu, 2011) utilizes concepts of mean-absolute deviation and Zadeh's extension principle to develop a solution method for the fuzzy portfolio optimization problem where the forecasted rate of return and risk are expressed as fuzzy numbers. It then constructs a pair of two-level mathematical programs to derive the upper bound and lower bound of the returns from the portfolio at a specific confidence level. The programs are then converted into a pair of ordinary one-level linear programs using duality theorem and variable transformation techniques. (Adachi & Takemura, 2011) use a bounded forecasting game formulated by Shafer and Vovk and neural network models to propose an investment strategy. (Chen et. al., 2011) shows the efficiency of the genetic relation algorithm with guided mutation to solve large scale portfolio problems. It adapts a genetic network programming based strategy for stock trading. (Dastkhan et. al., 2011) use a fuzzy weighted max-min mathematical model in a mean-absolute deviation portfolio selection problem with real features like minimum transaction lots, fixed and variable transaction costs, cardinality and bounds on holdings as constraints. To solve this problem, a hybrid genetic algorithm is proposed. (Pinto et. al., 2011) propose an autoregressive exogenous (ARX) predictor model to provide the risks and returns of securities in the Brazilian stock market. (Zhao et. al., 2011) propose a two-quadratic constrained data envelopment analysis (DEA) model for evaluation of mutual funds and considers risk and return as the two vital factors affecting mutual fund performance. (Zhu et. al., 2011) employ a PSO algorithm for portfolio selection and optimization. The model is tested on various unrestricted portfolios with no constraints on short selling and on restricted portfolios and then compared with a genetic algorithm (GA) to establish the superiority of the proposed PSO. (Zymlere et. al., 2011) combine robust portfolio optimization and classical portfolio insurance to hedge against events which are rare and not captured by an uncertainty set. The model enriches the portfolio with certain products in order to obtain a deterministic lower bound on the worst case return of the portfolio. (Asgharian, 2011) applies a conditional version of the optimal orthogonal portfolio (OOP) proposed by Mackinlay and Pastor by allowing for a timevarying risk premium. It captures the latent factors important for pricing a group of assets and constructs a portfolio representation of the OOP. The paper uses both a cross-sectional and a time series regression approach to evaluate the estimated OOP's out of sample pricing ability. (Daskalaki & Skiadopoulos, 2011) study the impact of including commodities in a portfolio with stocks, bonds and cash as asset classes. The paper finds that only when higher order moments are taken into account, commodities can offer in-sample diversification benefits in the portfolio. (Fortin & Hlouskova, 2011) examine the portfolio selection problem from an investor's point of view and not as a general equilibrium problem. The paper investigates how the maximization of a certain form of loss-averse (LA) utility relates to the optimization of CVaR. It formulates assumptions under which LA, MV and CVaR problems are equivalent. (Moon & Yao, 2011) propose a simple robust mean absolute deviation (MAD) model with reduced computational complexity and transforms the portfolio optimization problem into a linear programming problem. The proposed model considers parameter uncertainty by controlling the impact of estimation 5 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 errors on portfolio strategy performance. The study considers several factors like market condition, financial elasticity, standard deviation, parameter estimation period to specify conditions of superiority of the proposed robust optimization method over a nominal optimization method. (Garcia et. al., 2012) propose a re-sampling method based on time stamping mechanism, along with risk and return to address the uncertainty in these parameters during the optimization process. This approach is then tested with four popular MOEA for eight indices representing different asset classes with considerable reliability. (Huang, 2012) discusses a portfolio selection model in which Markowitz's MV problem is combined with uncertainty theory to select portfolios where returns are mainly given by estimations from experts. A method for calculating expected value, variance and semivariance is proposed which is integrated with GA to produce an intelligent algorithm. (Liu et. al., 2012) present two types of chance-variance (C-V) models where returns are characterized by fuzzy random variables with known possibility and probability distributions. The C-V models are formulated into equivalent stochastic programming which are solved by using a Monte-Carlo (MC) simulation based PSO algorithm. The MC simulated the probability distribution functions of the objective and constraints, while the PSO performs the search for optimal solutions. (Chen et. al., 2012) consider the long term asset class allocation problem and proposes a belief-rule-based (BRB) system which combines high dimensional return distributions of assets and portfolios to get the return distribution of a new portfolio. The BRB system solves non-linear portfolio optimization problem in the presence of non-linear cash flows and constraints. It then constructs the efficient frontier by using portfolios generated by BRB inference and then searches along the frontier to get the optimal portfolio. (Meskarian et. al., 2012) study the penalization scheme for stochastic programming models with SSD constraints. It applies this penalization scheme to some portfolio problems where the return functions of the assets are not always linear. (Li et. al., 2012) use credibility theory to characterize fuzziness and proposes a regret minimization model for fuzzy portfolio selection. The credibilistic model used in the paper leads to diversified investments unlike other credibilistic models which lead to concentrated investments. (Ballestero et. al., 2012) consider the case of environmentally responsible investments and deals with the problem of portfolio selection from the perspective of socially responsible investments. It classifies the assets as ethical and nonethical depending on the type of business the company is in and the transparency and credibility rating of the company. The investors are classified into three types based on the strength of their preference for ethical assets. (Patari et. al., 2012) discuss a portfolio-forming criteria based on the DEA efficiency scores for forming equity portfolios in which input and output factors are derived from stock valuation and price momentum indicators. (Nguyen & Lo, 2012) use a robust ranking model which is a minimax mixed integer programming problem for generating weights to maximize an objective function for the worst realization of the ranking. Constraints are generated using a network flow model. Ranking stocks itself gives a discrete uncertainty set. The authors then apply this method to portfolio optimization. (Zhang et. al., 2012) present the possibilistic mean-semivariance-entropy model for multi-period fuzzy portfolio selection. The authors consider four criteria: return, risk, transaction cost and diversification degree of the portfolio. The paper then proposes a hybrid intelligent algorithm which combines genetic algorithm (GA) and simulated annealing (SA) based techniques to solve the problem. (Tsaur, 2012) include a behavioral analysis of the investor's risk attitude to formulate a fuzzy portfolio model to solve the portfolio selection problem when return of assets are uncertain. (Fernandes et. al., 2012) propose the use of a portfolio optimization methodology which combines features of both Black Litterman (BL) methodology and re-sampling methodologies. It then generates a robust and diversified optimal allocation which are desirable to long term investors. (Hjalmarsson & Manchev, 2012) show how the weights in a MV optimization problem can be directly estimated as functions of the underlying stock characteristics, such as volume and momentum. It also studies the long-short portfolio choice in international MSCI indices for eighteen developed markets using three different 6 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 characteristics: book-to-market, dividend-price ratio and momentum. The paper brings forth the robust performance achieved by parameterizing portfolio weights directly as functions of underlying characteristics, unlike methods where the returns are modeled and estimated in an intermediate step. (Lebeault & Kharoubi, 2012) use an empirical copula analysis of the nonGaussian dependence structure on the equity market to study the diversification effect in portfolio optimization. It studies the impact of portfolio size on various risk measures over the period including the 2008 market crash. (Kourtis et. al., 2012) focus on the estimation of the inverse covariance matrix in the context of optimal portfolio choice using a 'shrinkage approach' to the maximum likelihood estimator of the inverse covariance matrix in order to replace portfolio risk and increase risk-adjusted returns. (Chen & Kwon, 2012) propose a robust selection problem for tracking a market index. The model is a 0-1 integer problem which avoids the computational difficulties of using quadratic tracking error by maximizing pair wise similarities between assets of the tracking portfolio and its target index. (Aranha et. al., 2012) extend the memetic tree-based genetic algorithm with the concept of terrain-based memetic algorithms for solving the portfolio optimization problem. (Huang & Qiao, 2012) introduce risk index as an alternate risk measure and employs uncertainty theory to solve a multi-period portfolio selection problem where the expected and standard deviation values of the uncertain security returns are given by expert's evaluations. (Yunusoglu & Selim, 2013) discuss the design of a fuzzy rule based expert system (ES) for portfolio managers. The ES considers the investor's risk profile and specific preferences by changing some parameters only. The ES outperforms all risk profiles when compared with a particular benchmark. (Tamiz et. al., 2013) incorporate macroeconomic, regional and country based factors in addition to factors specific to mutual funds into three variants of goal programming models for selection a portfolio of mutual funds across ten countries. (Cumning et. al., 2013) propose a modified appraisal value-based private equity (PE) benchmark. It shows that this method has statistically lower levels of risk than when listed PE indices are used as proxy. The listed PE indices are considered insufficient for portfolio optimization as they do not include the entire PE universe and their expected valuations often do not match the actually PE valuation, especially during crisis. (Miguel et. al., 2013) propose a new calibration criteria for shrinkage estimators of moments of asset returns and for shrinkage portfolios. It then studies a multivariate non-parametric approach to compute the optimal shrinkage intensity for independent and identically distributed returns. It also carried out a comprehensive empirical investigation of shrinkage estimators for portfolio selection on six empirical datasets. (Traglia & Gerlach, 2013) use a bivariate extreme value theory (EVT) model to study the extreme dependence of individual stocks and portfolios with the market index. It shows that lower tail dependence contains important information for risk averse investors and is relatively more stable over time than other risk measures like variance, semi-variance, skewness, beta, etc. (Behr et. al., 2013) devise a constrained minimum variance portfolio strategy to achieve portfolios with statistically significant higher Sharpe Ratio (SR). This strategy is built on the shrinkage estimation theory and imposes a data dependent structure on the empirical variance-covariance matrix estimate by trading off the reduction of sampling error and loss of sample information. This strategy achieves sizeable reductions in out-of-sample variances w.r.t. the other minimum variance portfolio strategies like constrained short selling, factor model, etc. (Li & Xu, 2013) propose a constrained multi-objective portfolio selection model with fuzzy random returns for investors. A compromise approach-based genetic algorithm is designed. The model has the capability to introduce judgment and expert opinion in accordance with the attitudes of the investors. (Gupta et. al., 2013) propose a multicriteria credibilistic portfolio selection fuzzy model integrated with a real-coded genetic algorithm (RCGA) which maximizes credibility such that the short-term return, long-term return and liquidity of the portfolio are greater than some given threshold levels. Several constraints, namely, capital budget constraint, cardinality constraint and diversification constraints are applied for investments in individual assets. (Lim et. al., 2013) develop a MV framework of portfolio selection based on 7 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 DEA cross-efficiency evaluation. The efficiency score and variance of the cross efficiencies of the decision making units (DMU) are used to represent the DMU's return and risk characteristics. Markowitz's MV formulation is then used to determine the inclusion of the DMU in a portfolio. (Takano & Gotoh, 2014) develop an efficient solution algorithm for kernel-based non-linear control policy based on dimensionality reduction technique. It then uses the technique for attaining high out-of-sample investment performance. (Utz et. al., 2014) discusses how inverse portfolio optimization can be used in a tri-criterion model. The third criteria causes the traditional non-dominated frontier to become a surface. Another empirical contribution is the application of the tri-criterion framework to socially responsible mutual funds. (Smimou, 2014) examines the impact of political instability on the composition of international portfolio under the discrete time version of Markowitz's MV portfolio selection problem. It studies to what extent can international diversification outperform domestic stock portfolios in presence of instability risk. (Levy & Levy, 2014) compare the performance of main optimization methods in literature when parameter estimation errors are accounted for. It then proposes two novel methods: variance-based constrained optimization (VBC) and global variance-based constrained optimization (GVBC) which take consider that estimation errors are larger for stocks with larger sample variance. (Castellano & Cerqueti, 2014) propose an extension of the MV Markowitz model by considering the presence of infrequently traded stocks and their effect on long-run optimal portfolio and shortterm trading strategies. (Bernand & Vanduffel, 2014) derive optimal portfolio with state dependent constraints by considering the dependence between the portfolio and the benchmark. The paper also derives tighter bounds on the Sharpe Ratio (SR) which is useful for fraud detection. (Fu et. al., 2014) study optimal asset allocation in a regime switching market comprising of an option, an underlying stock and a risk-free bond. The paper considers power and logarithmic utility functions and provides a solution to the portfolio optimization problem in incomplete markets. 3. Research Methodology In this section, we present a methodological framework for analyzing the issues and solutions regarding to portfolio selection. Initially a significant amount of literature was collected by searching through general databases such as 'Google Scholar' and 'EBSCO' using a combination of keywords like 'portfolio planning', 'financial portfolio selection', 'heuristic for portfolio optimization', etc from 2009 to 2014. It was assumed that most authors will definitely include the keyword 'portfolio' in the related articles. More than fifty-eight thousand results were available in Google Scholar which needed to be filtered. It included literature reviews, articles on project portfolio optimization, articles related to project portfolio management, book reviews, conference papers, articles in journals, etc. A cursory glance at few related abstracts revealed a large amount of repetition in the research materials when working papers and conference papers were later converted to journal papers, PhD and MS theses or reports led to publications in conferences and journals. So, in the first stage of the study, we had to decide which articles to consider and establish boundaries. To filter out irrelevant articles from the search results, it was decided to establish boundaries. This was done by considering to: (i) include only papers published in peer-reviewed journals. (ii) Papers were collected from five well-known journals to study the current trend of research in this field. (iii) The timeline was set from 2009 to 2014. Only papers which had theoretical, mathematical or empirical modeling were considered. Review papers were ignored. 8 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 Sl. No. 1. 2. 3. 4. 5. Table 2: Contribution of papers by journals for this review. Journal Title Papers Expert Systems with Applications European Journal of Operational Research Journal of Banking and Finance Computers & Operations Research Information Sciences 31 19 5 4 14 Percentage 42.47 26.03 6.85 5.48 19.18 Each paper was then selectively screened and finally seventy-three papers were considered for review. As the purpose of the study is to understand the current state of research and identify potential areas of concern in the portfolio management field, we try to find which author, country, year of publication, journal has the maximum contribution in this field and what are the performance measures and algorithms used to address this portfolio selection problem. We start by presenting the contribution of the selected journals to the field of portfolio management. This is listed in table-2. Next, we attempt to find out that as first authors only, which authors have contributed most to the recent literature. This is summarized in table-3. Sl. No. 1. 2. 3. Table 3: Number of papers as first author. First Author Number of papers as first author Yan Chen 3 Xiaoxia Huang 2 Kuang Yu Huang 2 Rest authors have contributed only once as first author. However, they may have contributed as second co-authors or above in other research papers which have been considered here, but that has not been analyzed. We now try to consider the number of authors in each of the publication and cluster them. This is shown in table-4. Table 4: Number of authors in each publication. Number of Number of Percentage authors publications 1 6 8.22 2 28 38.36 3 27 36.99 4 7 9.59 5 3 4.11 6 1 1.37 7 1 1.37 An important issue is the year of publication. Table-5 shows the number of publications in respective years. 9 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 Table 5: Number of publications in each year. Year of Number of papers Percentage Publication 2009 16 21.92% 2010 9 12.33% 2011 14 19.18% 2012 18 24.66% 2013 8 10.96% 2014 8 10.96% Another important issue is the nationality of the institutions that contributed to the literature. In order to obtain the relevant results, we kept a track of the nationality of the first authors. The coauthors (other than the first author) may have different nationality that the first author. This has not been considered during the calculation for the contribution of papers by various nations. The results are presented in table-6. Across these five journals, China and Taiwan dominate the contribution to the literature of 15.07% and 13.7% respectively. Table 6: Contribution of various countries in the current research. Country of First Number of Percentage Author papers Austria 1 1.37% Brazil 3 4.11% Canada 5 6.85% China 11 15.07% Finland 1 1.37% France 1 1.37% Germany 3 4.11% Greece 3 4.11% India 2 2.74% Iran 5 6.85% Israel 1 1.37% Italy 1 1.37% Japan 4 5.48% Korea 2 2.74% Kuwait 1 1.37% Norway 2 2.74% Singapore 1 1.37% Spain 4 5.48% Sweden 1 1.37% Taiwan 10 13.70% Turkey 2 2.74% UK 7 9.59% US 2 2.74% 10 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 Next, we study the use of performance attributes in the review papers. The table-7 shows the popular performance attributes and their frequency of occurrence in the publications. The analysis of table-7 clearly shows that Sharpe Ratio is the most commonly used performance attribute in portfolio management. Table 7: Performance measures and their occurrence in various papers Performance Measure Number of Percentage papers Alpha 3 4.48% Beta 1 1.49% Coefficient of Variation 1 1.49% CVaR 2 2.99% Entropy 1 1.49% Farinelli-Tibiletti Herfindahl 1 1.49% Index Financial Ratios 1 1.49% Hypervolume 1 1.49% Information Ratio 3 4.48% Intraclass Inertia 1 1.49% M-2 2 2.99% MAPE 1 1.49% Market Ratio 1 1.49% Mean Runtime 3 4.48% Moments 1 1.49% Omega 1 1.49% Portfolio Turnover 3 4.48% Profitability 3 4.48% Return on Investment 1 1.49% Sharpe Ratio 25 37.31% Skewness 1 1.49% Sortino Ratio 1 1.49% Spearman's Rank 1 1.49% Correlation Coefficient Tracking Error 1 1.49% Training Error 1 1.49% Treynor Ratio 3 4.48% Turnover Rate 1 1.49% VaR 2 2.99% Next, we can check how many performance measures have been used together in a single publication. The table-8 shows this. 11 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 Table 8: Number of performance measures in each paper. Number of performance Number of Percentage measures publications 0 24 33.80 1 33 46.48 2 9 12.68 3 2 2.82 4 2 2.82 6 1 1.41 Clearly, most papers use one performance measure and generally it is Sharpe Ratio (Wang & Huang, 2011; Zhu et.al., 2011; Smimou, 2014; Levy & Levy, 2014; Castellano & Cerqueti, 2014; Bernard & Vanduffel, 2014; Lim et.al., 2013; Asgharian, 2011; Zakamouline & Koekebakkar, 2009; Alexander & Baptista, 2010; Hjalmarsson & Manchev, 2012; Cumming et.al., 2013; DeMiguel et.al., 2013; Behr et.al., 2013; Aranha et.al., 2012; Patari et.al. 2012; Barros et.al., 2012; Diaz et.al., 2009; Daskalaki & Skiadoupoulos, 2011; Kourtis et.al., 2012; Zymler et.al., 2011; Fortim & Hlouskova, 2011; Yunusoglu & Selim, 2013; Zhao et.al., 2011). In spite of the drawbacks of the Sharpe Ratio, its frequent usage may be because of the ease of calculation. Another important area of concern is the study of the various algorithms used in the field of portfolio management. The table-9 shows the area of focus of the algorithms used to address the portfolio selection problem. 12 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 Table 9: Various algorithms used in the field of portfolio management. Algorithm Number of papers Analytic Hierarchy Process 1 Adaptive Resonance Theory 1 Autoregressive Exogenous 3 Prediction Back propagation Neural 1 Network Belief Rule based System 1 Cluster Analysis 1 Constrained Optimization 1 models Data Envelopment Analysis 3 Evolutionary Algorithm 2 Entropy 2 Expert Systems methodology 3 Extreme Value Theory 1 Field Theory 1 Fuzzy Logic 13 Genetic Algorithm variants 11 Game Theory 1 Genetic Network Programming 3 Goal Programming 2 Grey System 2 K-means Clustering 1 Mathematical Modeling 17 Memetic Algorithm 1 Mixed Integer Programming 1 Model Multiobjective Evolutionary 2 Algorithm Multi Attribute Decision 1 Analysis Neural Network 2 Possibilistic Theory 1 Particle Swarm Optimization 4 Quadratic Programming Model 3 Resampling Methodology 1 Robust Portfolio Optimization 2 Rough Set 2 Self Organizing Maps 1 Simulated Annealing 1 Stochastic Programming 1 Models 13 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 A final issue that we will address is the classification of the papers according to the methodological approach followed by the authors. For carrying this, we went through the titles and abstracts and when that was inconclusive, we read through the main body of the paper. Reviewing the papers, we find that these are the main methods followed in the papers which are listed in table-10. Table 10: Methodology followed by the authors Methodological Description Number of Category papers Empirical Experiment Papers focusing on qualitative 4 or quantitative analysis of data. Theoretical Papers proposing a 2 framework without any testing with real data. Model Design Papers developing a new 3 algorithm or model to explain the problem. Combined Papers using a combination 64 of either of the above. Percentage 5.48 2.74 4.11 87.67 We can see that most papers use a combination of two or more methods which is pretty natural considering that the authors would like to validate their proposed framework or model with stock market data. 5. Conclusion The purpose of this article is to provide an insight into the current state of research in the area of portfolio management. In order to understand this, we have analyzed seventy-three shortlisted papers from five major journals from 2009-2014 by classifying them on various headers. Our analysis revealed that only 8.22 percent of the papers are single authored (Huang, 2009; Liu, 2011; Huang, 2012; Tsaur, 2012; Smimou, 2014; Asgharian, 2011). Rest all papers are coauthored. We have analyzed the contribution of the various nationalities in the research in portfolio management. Sharpe Ratio is the most prominently used performance measure with 37.31 percent of the papers using it. In spite of the shortcomings of Sharpe Ratio, this may be because of the ease of calculation of this ratio compared to other performance attributes. Most papers use a single performance measure which is the Sharpe Ratio. On analyzing the broader area of the algorithms used to address the problem of problem management, it was found that fuzzy logic, genetic algorithm and its variants like genetic network programming, combined genetic algorithm, etc. have been used in many papers in addition to mathematical modeling. A final issue concerns the methodological choice of the authors in their papers. About 87.67 percent authors use a combination of any two of empirical-experiment, theoretical and model designing approach. 6. Future Research The decision making process in portfolio optimization is a very complex problem and should consider practical constraints like cardinality constraints, transaction costs, sector capitalization constraints. Research can be done to check how the models have evolved incorporating these constraints. Also, we have not checked the number of objective functions and risk measures 14 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 used by the authors in their papers. This can shed light on the multi-objective algorithms being designed to solve complex problems like the portfolio optimization problem. We find maximum use of Sharpe Ratio as a performance attribute in spite of its limitations. Better performance indicators should be used to assess the performance of the models. We have tried to identify some of the possible directions of future research in the portfolio management field. It is evident that this field presents intense challenges and significant research opportunities for interested academicians and practitioners. 7. References Adachi, R., & Takemura, A. 2011. Sequential optimizing investing strategy with neural networks. Expert Systems with Applications, 38(10), 12991-12998. Alexander, G. J., & Baptista, A. M. 2010. Active portfolio management with benchmarking: A frontier based on alpha. Journal of Banking & Finance, 34(9), 2185-2197. Alexander, S., Coleman, T. F., & Li, Y. 2006. Minimizing CVaR and VaR for a portfolio of derivatives. Journal of Banking & Finance, 30(2), 583-605. Anagnostopoulos, K. P., & Mamanis, G. 2010. A portfolio optimization model with three objectives and discrete variables. Computers & Operations Research,37(7), 1285-1297. Aranha, C., Azevedo, C. R., & Iba, H. 2012. Money in trees: How memes, trees, and isolation can optimize financial portfolios. Information Sciences,182(1), 184-198. Asgharian, H. 2011. A conditional asset-pricing model with portfolio. Journal of Banking & Finance, 35(5), 1027-1040. the optimal orthogonal Ballestero, E., Bravo, M., Pérez-Gladish, B., Arenas-Parra, M., & Plà-Santamaria, D. 2012. Socially Responsible Investment: A multicriteria approach to portfolio selection combining ethical and financial objectives.European Journal of Operational Research, 216(2), 487-494. Barros Fernandes, J. L., Haas Ornelas, J. R., & Martínez Cusicanqui, O. A. 2012. Combining equilibrium, resampling, and analyst’s views in portfolio optimization. Journal of Banking & Finance, 36(5), 1354-1361. Behr, P., Guettler, A., & Miebs, F. 2013. On portfolio optimization: imposing the right constraints. Journal of Banking & Finance, 37(4), 1232-1242. Bernard, C., & Vanduffel, S. 2014. Mean–variance optimal portfolios in the presence of a benchmark with applications to fraud detection. European Journal of Operational Research, 234(2), 469-480. Bey, R. P., & Howe, K. M. 1984. Gini's mean difference and portfolio selection: An empirical evaluation. Journal of Financial and Quantitative Analysis, 19(03), 329-338. Branke, J., Scheckenbach, B., Stein, M., Deb, K., & Schmeck, H. (2009). Portfolio optimization with an envelope-based multi-objective evolutionary algorithm. European Journal of Operational Research, 199(3), 684-693. 15 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 Byrne, P., & Lee, S. 2004. Different risk measures: different portfolio compositions?. Journal of Property Investment & Finance, 22(6), 501-511. Castellano, R., & Cerqueti, R. 2014. Mean–Variance portfolio selection in presence of infrequently traded stocks. European Journal of Operational Research, 234(2), 442-449. Chang, C. H., Lin, J. J., Lin, J. H., & Chiang, M. C. 2010. Domestic open-end equity mutual fund performance evaluation using extended TOPSIS method with different distance approaches. Expert systems with applications, 37(6), 4642-4649. Chang, T. J., Yang, S. C., & Chang, K. J. 2009. Portfolio optimization problems in different risk measures using genetic algorithm. Expert Systems with Applications, 36(7), 10529-10537. Chekhlov, A. V., Uryasev, S., & Zabarankin, M. 2000. Portfolio optimization with drawdown constraints. Department of Industrial & Systems Engineering, University of Florida. Chen, C., & Kwon, R. H. 2012. Robust portfolio selection for index tracking.Computers & Operations Research, 39(4), 829-837. Chen, J. S., & Lin, Y. T. 2009. A partitioned portfolio insurance strategy by a relational genetic algorithm. Expert Systems with Applications, 36(2), 2727-2734. Chen, J. S., Hou, J. L., Wu, S. M., & Chang-Chien, Y. W. (2009). Constructing investment strategy portfolios by combination genetic algorithms. Expert Systems with Applications, 36(2), 3824-3828. Chen, L. H., & Huang, L. 2009. Portfolio optimization of equity mutual funds with fuzzy return rates and risks. Expert Systems with Applications, 36(2), 3720-3727. Chen, Y. W., Poon, S. H., Yang, J. B., Xu, D. L., Zhang, D., & Acomb, S. 2012. Belief rule-based system for portfolio optimisation with nonlinear cash-flows and constraints. European Journal of Operational Research, 223(3), 775-784. Chen, Y., Mabu, S., & Hirasawa, K. 2010. A model of portfolio optimization using time adapting genetic network programming. Computers & operations research, 37(10), 1697-1707. Chen, Y., Mabu, S., & Hirasawa, K. 2011. Genetic relation algorithm with guided mutation for the large-scale portfolio optimization. Expert Systems with Applications, 38(4), 3353-3363. Chen, Y., Ohkawa, E., Mabu, S., Shimada, K., & Hirasawa, K. 2009. A portfolio optimization model using Genetic Network Programming with control nodes. Expert Systems with Applications, 36(7), 10735-10745. Chiam, S. C., Tan, K. C., & Mamun, A. A. 2009. A memetic model of evolutionary PSO for computational finance applications. Expert Systems with Applications, 36(2), 3695-3711. Cumming, D., Helge Haß, L., & Schweizer, D. 2013. Private equity benchmarks and portfolio optimization. Journal of Banking & Finance, 37(9), 3515-3528. 16 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 Daskalaki, C., & Skiadopoulos, G. 2011. Should investors include commodities in their portfolios after all? New evidence. Journal of Banking & Finance, 35(10), 2606-2626. Dastkhan, H., Gharneh, N. S., & Golmakani, H. 2011. A linguistic-based portfolio selection model using weighted max–min operator and hybrid genetic algorithm. Expert Systems with Applications, 38(9), 11735-11743. DeMiguel, V., Martin-Utrera, A., & Nogales, F. J. 2013. Size matters: Optimal calibration of shrinkage estimators for portfolio selection. Journal of Banking & Finance, 37(8), 3018-3034. Desmoulins-Lebeault, F., & Kharoubi-Rakotomalala, C. 2012. Non-Gaussian diversification: When size matters. Journal of Banking & Finance, 36(7), 1987-1996. Díaz, A., González, M. D. L. O., Navarro, E., & Skinner, F. S. 2009. An evaluation of contingent immunization. Journal of Banking & Finance, 33(10), 1874-1883. DiTraglia, F. J., & Gerlach, J. R. 2013. Portfolio selection: An extreme value approach. Journal of Banking & Finance, 37(2), 305-323. Elton, E. J., & Gruber, M. J. 1974. On the optimality of some multiperiod portfolio selection criteria. Journal of Business, 231-243. Elton, E. J., & Gruber, M. J. 1997. Modern portfolio theory, 1950 to date.Journal of Banking & Finance, 21(11), 1743-1759. Esfahanipour, A., & Aghamiri, W. 2010. Adapted neuro-fuzzy inference system on indirect approach TSK fuzzy rule base for stock market analysis.Expert Systems with Applications, 37(7), 4742-4748. Fan, A., & Palaniswami, M. 2001. Stock selection using support vector machines. In Neural Networks, 2001. Proceedings. IJCNN'01. International Joint Conference on (Vol. 3, pp. 1793-1798). IEEE. Fasanghari, M., & Montazer, G. A. 2010. Design and implementation of fuzzy expert system for Tehran Stock Exchange portfolio recommendation. Expert Systems with Applications, 37(9), 6138-6147. Fortin, I., & Hlouskova, J. 2011. Optimal asset allocation under linear loss aversion. Journal of Banking & Finance, 35(11), 2974-2990. Fu, J., Wei, J., & Yang, H. 2014. Portfolio optimization in a regime-switching market with derivatives. European Journal of Operational Research, 233(1), 184-192. Gaivoronski, A. A., & Pflug, G. 2005. Value-at-risk in portfolio optimization: properties and computational approach. Journal of Risk, 7(2), 1-31. García, S., Quintana, D., Galván, I. M., & Isasi, P. 2012. Time-stamped resampling for robust evolutionary portfolio optimization. Expert Systems with Applications, 39(12), 10722-10730. 17 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 Gil-Aluja, J., Gil-Lafuente, A. M., & Merigó, J. M. 2011. Using homogeneous groupings in portfolio management. Expert Systems with Applications, 38(9), 10950-10958. Golmakani, H. R., & Fazel, M. 2011. Constrained portfolio selection using particle swarm optimization. Expert Systems with Applications, 38(7), 8327-8335. Gupta, P., Inuiguchi, M., Mehlawat, M. K., & Mittal, G. 2013. Multiobjective credibilistic portfolio selection model with fuzzy chance-constraints. Information Sciences, 229, 1-17. Hakansson, N. H. 1971. MULTI‐PERIOD MEAN‐VARIANCE ANALYSIS: TOWARD A GENERAL THEORY OF PORTFOLIO CHOICE*. The Journal of Finance, 26(4), 857-884. Hart, C. 1998. Doing a literature review: Releasing the social science research imagination. Sage. Hjalmarsson, E., & Manchev, P. 2012. Characteristic-based choice. Journal of Banking & Finance, 36(5), 1392-1401. mean-variance portfolio Huang, K. Y. 2009. Application of VPRS model with enhanced threshold parameter selection mechanism to automatic stock market forecasting and portfolio selection. Expert Systems with Applications, 36(9), 11652-11661. Huang, K. Y., & Jane, C. J. 2009. A hybrid model for stock market forecasting and portfolio selection based on ARX, grey system and RS theories. Expert Systems with Applications, 36(3), 5387-5392. Huang, X. 2009. A review of credibilistic portfolio selection. Fuzzy Optimization and Decision Making, 8(3), 263-281. Huang, X. 2012. Mean–variance models for portfolio selection estimations. Expert Systems with Applications, 39(5), 5887-5893. subject to experts’ Huang, X., & Qiao, L. 2012. A risk index model for multi-period uncertain portfolio selection. Information Sciences, 217, 108-116. Inuiguchi, M., & Ramık, J. 2000. Possibilistic linear programming: a brief review of fuzzy mathematical programming and a comparison with stochastic programming in portfolio selection problem. Fuzzy sets and systems, 111(1), 3-28. Jensen, M. C. 1968. The performance of mutual funds in the period 1945–1964. The Journal of finance, 23(2), 389-416. Kendall, G., & Su, Y. 2005. A Particle Swarm Optimisation Approach in the Construction of Optimal Risky Portfolios. In Artificial Intelligence and Applications (pp. 140-145). Konno, H., Shirakawa, H., & Yamazaki, H. 1993. A mean-absolute deviation-skewness portfolio optimization model. Annals of Operations Research, 45(1), 205-220. Kourtis, A., Dotsis, G., & Markellos, R. N. 2012. Parameter uncertainty in portfolio selection: Shrinking the inverse covariance matrix. Journal of Banking & Finance, 36(9), 2522-2531. 18 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 Levy, H., & Levy, M. 2014. The benefits of differential variance-based constraints in portfolio optimization. European Journal of Operational Research,234(2), 372-381. Li, D., & Ng, W. L. 2000. Optimal Dynamic Portfolio Selection: Multiperiod Mean‐Variance Formulation. Mathematical Finance, 10(3), 387-406. Li, J., & Xu, J. 2013. Multi-objective portfolio selection model with fuzzy random returns and a compromise approach-based genetic algorithm. Information Sciences, 220, 507-521. Li, X., Shou, B., & Qin, Z. 2012. An expected regret minimization portfolio selection model. European Journal of Operational Research, 218(2), 484-492. Lim, S., Oh, K. W., & Zhu, J. 2013. Use of DEA cross-efficiency evaluation in portfolio selection: An application to Korean stock market. European Journal of Operational Research. Lin, P. C., & Ko, P. C. 2009. Portfolio value-at-risk forecasting with GA-based extreme value theory. Expert Systems with Applications, 36(2), 2503-2512. Lintner, J. 1965. The valuation of risk assets and the selection of risky investments in stock portfolios and capital budgets. The review of economics and statistics, 13-37. Liu, S. T. 2011. A fuzzy modeling for fuzzy portfolio optimization. Expert Systems with Applications, 38(11), 13803-13809. Liu, Y., Wu, X., & Hao, F. 2012. A new Chance–Variance optimization criterion for portfolio selection in uncertain decision systems. Expert Systems with Applications, 39(7), 65146526. Loraschi, A., Tomassini, M., Tettamanzi, A., & Verda, P. 1995. Distributed genetic algorithms with an application to portfolio selection problems. In Artificial neural nets and genetic algorithms (pp. 384-387). Springer Vienna. Markowitz, H. 1952. Portfolio selection*. The journal of finance, 7(1), 77-91. Markowitz, Harry. 1959. Portfolio Selection: Efficient Diversification of Investments. Cowles Foundation Monograph No. 16. New York: John Wiley &Sons, Inc. Meskarian, R., Xu, H., & Fliege, J. 2012. Numerical methods for stochastic programs with second order dominance constraints with applications to portfolio optimization. European Journal of Operational Research, 216(2), 376-385. Metaxiotis, K., & Liagkouras, K. 2012. Multiobjective Evolutionary algorithms for portfolio management: a comprehensive literature review. Expert Systems with Applications, 39(14), 11685-11698. Moon, Y., & Yao, T. 2011. A robust mean absolute deviation model for portfolio optimization. Computers & Operations Research, 38(9), 1251-1258. Mossin, J. 1968. Optimal multiperiod portfolio policies. Journal of Business,41(2), 215. 19 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 Nanda, S. R., Mahanty, B., & Tiwari, M. K. 2010. Clustering Indian stock market data for portfolio management. Expert Systems with Applications,37(12), 8793-8798. Nguyen, T. D., & Lo, A. W. 2012. Robust ranking and portfolio optimization. European Journal of Operational Research, 221(2), 407-416. Pätäri, E., Leivo, T., & Honkapuro, S. 2012. Enhancement of equity portfolio performance using data envelopment analysis. European Journal of Operational Research, 220(3), 786-797. Philippatos, G. C., & Wilson, C. J. 1972. Entropy, market risk, and the selection of efficient portfolios. Applied Economics, 4(3), 209-220. Pinto, D. D. D., Monteiro, J. G. M. S., & Nakao, E. H. 2011. An approach to portfolio selection using an ARX predictor for securities’ risk and return. Expert Systems with Applications, 38(12), 15009-15013. Price, K., Price, B., & Nantell, T. J. 1982. Variance and lower partial moment measures of systematic risk: some analytical and empirical results. The Journal of Finance, 37(3), 843855. Rockafellar, R. T., & Uryasev, S. 2002. Conditional value-at-risk for general loss distributions. Journal of Banking & Finance, 26(7), 1443-1471. Rödder, W., Gartner, I. R., & Rudolph, S. 2010. An entropy-driven expert system shell applied to portfolio selection. Expert Systems with Applications,37(12), 7509-7520. Samuelson, P. 1969. Lifetime portfolio selection by dynamic stochastic programming. Shalit, H., & Yitzhaki, S. 1984. Mean‐Gini, Portfolio Theory, and the Pricing of Risky Assets. The Journal of Finance, 39(5), 1449-1468. Sharpe, W. F. 1964. Capital asset prices: A theory of market equilibrium under conditions of risk*. The journal of finance, 19(3), 425-442. Sharpe, W. F. 1966. Mutual fund performance. Journal of business, 119-138. Sharpe, W. F. 1967. A linear programming selection. Management Science, 13(7), 499-510. algorithm for mutual fund portfolio Shaw, J. 1995. A schema approach to the formal literature review in engineering theses. System, 23(3), 325-335. Smimou, K. 2014. International portfolio choice and political instability risk: A multi-objective approach. European Journal of Operational Research, 234(2), 546-560. Soleimani, H., Golmakani, H. R., & Salimi, M. H. (2009). Markowitz-based portfolio selection with minimum transaction lots, cardinality constraints and regarding sector capitalization using genetic algorithm. Expert Systems with Applications, 36(3), 5058-5063. 20 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 Takano, Y., & Gotoh, J. Y. 2014. Multi-period portfolio selection using kernel-based control policy with dimensionality reduction. Expert Systems with Applications, 41(8), 3901-3914. Tamiz, M., Azmi, R. A., & Jones, D. F. 2013. On selecting portfolio of international mutual funds using goal programming with extended factors. European Journal of Operational Research, 226(3), 560-576. Tanaka, H., Guo, P., & Türksen, I. B. 2000. Portfolio selection based on fuzzy probabilities and possibility distributions. Fuzzy sets and systems, 111(3), 387-397. Tiryaki, F., & Ahlatcioglu, B. 2009. Fuzzy portfolio selection using fuzzy analytic hierarchy process. Information Sciences, 179(1), 53-69. Treynor, J. L. 1965. How to rate management of investment funds. Harvard business review, 43(1), 63-75. Tsaur, R. C. 2012. Fuzzy portfolio model with different investor risk attitudes.European Journal of Operational Research. Utz, S., Wimmer, M., Hirschberger, M., & Steuer, R. E. 2014. Tri-criterion inverse portfolio optimization with application to socially responsible mutual funds. European Journal of Operational Research, 234(2), 491-498. Wang, K., & Huang, S. 2010. Using fast adaptive neural network classifier for mutual fund performance evaluation. Expert Systems with Applications, 37(8), 6007-6011. Webster, J., & Watson, R. T. 2002. Analyzing the past to prepare. MIS quarterly, 26(2), 13-23. Xia, Y., Liu, B., Wang, S., & Lai, K. K. 2000. A model for portfolio selection with order of expected returns. Computers & Operations Research, 27(5), 409-422. Xidonas, P., Ergazakis, E., Ergazakis, K., Metaxiotis, K., Askounis, D., Mavrotas, G., & Psarras, J. 2009. On the selection of equity securities: An expert systems methodology and an application on the Athens Stock Exchange. Expert Systems with Applications, 36(9), 1196611980. Yunusoglu, M. G., & Selim, H. 2013. A fuzzy rule based expert system for stock evaluation and portfolio construction: An application to Istanbul Stock Exchange. Expert Systems with Applications, 40(3), 908-920. Zakamouline, V., & Koekebakker, S. 2009. Portfolio performance evaluation with generalized Sharpe ratios: Beyond the mean and variance. Journal of Banking & Finance, 33(7), 12421254. Zenios, S. A., & Kang, P. 1993. Mean-absolute deviation portfolio optimization for mortgagebacked securities. Annals of Operations Research, 45(1), 433-450. Zhang, W. G., Liu, Y. J., & Xu, W. J. 2012. A possibilistic mean-semivariance-entropy model for multi-period portfolio selection with transaction costs. European Journal of Operational Research, 222(2), 341-349. 21 Proceedings of 10th Global Business and Social Science Research Conference 23 -24 June 2014, Radisson Blu Hotel, Beijing, China, ISBN: 978-1-922069-55-9 Zhang, W. G., Zhang, X. L., & Xiao, W. L. (2009). Portfolio selection under possibilistic mean– variance utility and a SMO algorithm. European Journal of Operational Research, 197(2), 693-700. Zhao, X., Wang, S., & Lai, K. K. 2011. Mutual funds performance evaluation based on endogenous benchmarks. Expert Systems with Applications, 38(4), 3663-3670. Zhu, H., Wang, Y., Wang, K., & Chen, Y. (2011). Particle Swarm Optimization (PSO) for the constrained portfolio optimization problem. Expert Systems with Applications, 38(8), 1016110169. Zymler, S., Rustem, B., & Kuhn, D. 2011. Robust portfolio optimization with derivative insurance guarantees. European Journal of Operational Research,210(2), 410-424. 22