PORTFOLIO MANAGEMENT OF INSTITUTIONAL INVESTORS IN A DECISION SUPPORT SYSTEMS FRAMEWORK
by
M E SRINIVAS
Department of Management Studies
THESIS SUBMITTED
IN FULFILMENT OF THE REQUIREMENTS FOR THE AWARD OF THE DEGREE OF
DOCTOR OF PHILOSOPHY
to the
INDIAN INSTITUTE OF TECHNOLOGY, DELHI
INDIA
DECEMBER, 1993
(Dr. P.K JAIN)
".7 (Dr. SUSHIL)
CERTIFICATE
The thesis entitled "PORTFOLIO MANAGEMENT OF INSTITUTIONAL INVESTORS IN A DECISION SUPPORT SYSTEMS FRAMEWORK", being submitted by Mr.M.E.SRINIVAS to the Indian Institute of Technology, Delhi, for the award of the degree of DOCTOR OF PHILOSOPHY, is a record of bonafide research work carried out by him. He has worked under our guidance and supervision, and has fulfilled the requirements for the submission of this thesis which has attained the standard required for a Ph.D degree of the Institute. The results presented in this thesis have not been submitted elsewhere for the award of any degree or diploma.
Professor and Head, Associate Professor, Department of Management Studies Department of Management Studies Indian Institute of Technology, Indian Institute of Technology,
New Delhi. New Delhi.
2,edecated .6* oaf Pevteard
ACKNOWLEDGEMENTS
I express my deep sense of gratitude to Prof.P.K.Jain and Dr.Sushil. T. am extremely fortunate to have had them as my research supervisors. Ever patient and accessible,Prof.P.K.Jain devoted considerable time to guiding my research study despite his heavy official responsibilities. Dr.Sushil has provided able guidance and inspiring encouragement throughout the study.
The study has benefitted from the suggestions and valuable inputs from several persons and organizations. I am greatly indebted to all of them. I am especially grateful to Prof.M.Tiripalraju,Vice Principal,UTI Institute of Capital Markets, who, with his keen interest in the study, was kind enough to extend all possible help during my visit to Bombay. The study has gained immensely from his comments and helpful suggestions. I must put on record my sincere thanks to Mr.Sanjeev Pandiya of SRF Finance Ltd.,Mr.Pawan K.Chaudhary of Small Industries Development Bank of India (SIDBI), besides the top management personnel of the various mutual fund organizations and investment firms, without whose active help and cooperation, the study would not have materialized.
acknowledge with gratitude the help extended by the learned faculty of the Department of Management Studies. I am particularly thankful to Prof.Prem Vrat and Prof.Vinayshil Gautam,ex-heads of the department, for supporting the study. I must also thank the staff and friends at the department for their cooperation.
Thanks are due to Prof. H.C.Jain,Librarian,Delhi University
(South Campus),for granting me liberal access to the library; to the Head and staff of the computer service centre,IIT Delhi;and to the Librarian,Delhi School of Economics,for their kind help.
Finally,I express my appreciation and gratitude to my friends for the invaluable help they provided at various stages of the research study. In particular, I am extremely grateful to Mr.Girish Morey, Dr.P.V.Siva Rama Prasad, Mr.D.Srinivas, Mr.Kota Kameswara Rao, Mr. K.Ramachandran, Mr. B.V.V.S.N.Prabhakara Rao, Mr. T. V. B.
P. S. Rama Krishna, Mr. Roy Abraham, Mr.Sai K.Reddy, Col.T.N.Baba, Mr.G.Srinivas, Mr.Paila Nagesh, and Mr.P.T.Shriprasad.
j
taj,..-imu\ak
(M.E.Srinivas)
ABSTRACT
Portfolio management function is admittedly a complex task.
The problem is innately dynamic, involving several alternatives, multiple goals and criteria, with a certain amount of uncertainty surrounding it. In effect, therefore, the portfolio management problem may at best be described as a semi-structured one. The objective of the present study is to develop a Decision Support Systems framework to aid the decision making process of the portfolio managers of institutional investors. In developing such a framework, the thesis addresses itself to the issues of complex- ity, dynamics, multiple criteria associated with the problem, the need for flexible systems methodology, and qualitative and quantitative modelling for portfolio management.
The first phase of the study involved identification of the problem, followed by a detailed survey of literature related to the problem classified under the following subject areas, viz., Modelling for portfolio analysis, Asset pricing and valuation, Market behaviour, Practical issues in portfolio management, Decision Support Systems and Modelling Methodologies. A survey of the portfolio investment practices of professional managers in the Indian context was also carried out to gain insight into practical issues. The survey as well as the literature review enabled the conceptualization of the framework of Decision Support Systems for portfolio management. The DSS is envisaged to consist of three subsystems, viz., Database management system, Model base management system and Dialogue management system. The following set of models are identified for the purpose : Analytic Hierarchy Process (AHP) model, Fuzzy Set Outranking methodology model, System Dynamics (SD) model, Goal Programming (GP) model, and Fuzzy Goal Programming (FGP) model. The study attempts to identify some of the important data elements that have a bearing on the portfolio management decision making. Sample data is collected on a selected list of securities for the period 1980-90, and their beta parameters estimated.
The AHP model grapples with the complexity involved by breaking down the problem into a hierarchical structure, consisting of four levels, viz., investor's satisfaction, the objectives for portfolio management, the criteria for the objectives, and finally, the industry groups or the securities considered for inclusion in the portfolio. The Fuzzy Set Outranking model also follows similar pattern. This methodology allows impreciseness in expert judgements while rating industries/securities on various features. Both these models are applied in the Indian context.
The thesis develops a System Dynamics model for simulation of security prices and illustrates its application to portfolio management. The SD model is validated in case of one security and applied to the securities comprising the Bombay Stock Exchange Sensitive Index for simulation of their year-end prices. The GP model formulated in the study is a quantitative model, applicable when investor's goals can be clearly stated, while the FGP model allows ambiguity in goal specification. In the study, the GP and FGP models have been applied in the Indian context.
The Thesis synthesizes the DSS framework, specifying some of its important features in terms of inputs, outputs and the what-if and what-to-do-if capabilities. The issues involved in implementa- tion of the framework are also highlighted.Software for application of the models is developed during the course of the research work.
(iv
)CONTENTS
Page No.
ABSTRACT (iii)
LIST OF FIGURES (x)
LIST OF TABLES (xii)
LIST OF APPENDICES (xiv)
NOMENCLATURE (xv)
CHAPTER 1.1 1.2
I INTRODUCTION TO THE STUDY Introduction
Problem Definition 1.2.1 Objectives 1.2.2 Issues
1.2.3 Scope of the study
1
3 3 4 5
1.3 Methodology of the study 6
1.4 Organisation of the Report 7
1.5 Concluding Remarks 11
CHAPTER 2 LITERATURE REVIEW
2.1 Introduction 12
2.2 Basis of the present Review 12 2.3 Modelling for Portfolio Analysis 13 2.4 Asset Pricing and Valuation 23 2.4.1 Capital Asset Pricing Model 26
2.4.2 Studies on Beta 27
2.4.3 The Validity of the CAPM 29 2.4.4 Other Asset Pricing Approaches 33 2.4.5 Other Considerations Regarding Risk 34 2.4.6 Other Valuation Attributes 36
2.5 Market Behaviour 38 2.5.1 Market Efficiency 38
2.5,2 Market Anomalies 41
2.5.3 Other Studies 45
2.5.4 Indian Studies 47
2.5.5 New Perspectives 49
2.6 Portfolio Investment Practices 54 2.7 Decision Support Systems 66 2.8 Modelling Methodologies 73 2.8.1 Analytic Hierarchy Process 74 2.8.2 Fuzzy Set Outranking Methodology 75
2.8.3 Goal Programming 75
2.8.4 Fuzzy Goal Programming 76
2.8.5 System Dynamics 76
2.9 The Current State of the Research 78 2.10 Limitations of the Existing Approaches 83 2.11 Need for Further Research and Areas of study 85
2.12 Concluding Remarks 86
CHAPTER 3 PRESENT PRACTICES AND AN OVERVIEW OF THE PROPOSED DSS FRAMEWORK
3.1 Introduction 87
3.2 Survey of Practices 87
3.2.1 Survey Methodology 87 3.2.2 Important Findings of the Survey 89 3.3 Overview of the Proposed DSS Framework 93 3.3.1 Conceptual Framework 93 3.3.2 Relevance in Different Decision making phase896
3.3.3 The Model Base 98
3.4 Concluding Remarks 103
CHAPTER 4 SAMPLE DATA FOR THE DATABASE 105
4.1 Introduction 105
4.2 Database for the DSS
4.2.1 Database System Concepts 105 4.2.2 Suggested Approach for Conceptual View of
Data 108
4.2.3 Database Elements for the DSS 110 4.3 Sample Data for the Study 115
4.3.1 Data Collected 115
4.3.2 Estimation of the Beta parameter 117 4.4 Use of Database in the Models 136
4.5 Concluding Remarks 138
CHAPTER 5 ANALYTIC HIERARCHY PROCESS MODEL
5.1 Introduction 140
5.2 Analytic Hierarchy Process (AHP)-An Overview 141 5.2.1 Principles Involved 141 5.2.2 Methodology
5.3 Outranking Problems in Portfolio Management 144
5.4 Proposed Model 146
5.4.1 The Industries Model 146 5.4.2 The securities Model 151
5.5 Case Study 154
5.5.1 Industries Model 155 5.5.2 Securities Model 159
5.6 Concluding Remarks 176
CHAPTER 6 FUZZY SET OUTRANKING MODEL
6.1 Introduction 178
6.2 Need for Fuzzy Set Methodology for Portfolio
Management 178
6.3 Theoretical Background 179
6.3.1 Fuzzy Sets 179 6.3.2 Fuzzy Set Outranking Methodology 180 6.4 Portfolio Allocation Model using Fuzzy set methodology 182 6.4.5 Industries Model 182 6.4.2 Securities Model 184
6.5 Case Study 185
6.5.1 Industries Model 185 6.5.2 Securities Model 191
6.6 Concluding Remarks 206
CHAPTER 7 GOAL PROGRAMMING AND SYSTEM DYNAMICS MODELS
7.1 Introduction - 207
7.2 Need for Goal Programming in Conjunction with System Dynamics for Portfolio Management 207 7.2.1 Need for Goal Programming (GP) 207 7.2.2 Need for system Dynamics (SD) 208 7.3 Theoretical Background 209 7.3.1 Goal Programming 209
7.3.2 System Dynamics 212
7.4 Portfolio Management Using Goal Programming and System
Dynamics 216
7.4.1 The GP Model 216
7.4.2 SD Model for Estimation of Expected Returns for Individual Securities 228
7.5 Case Study 238
7.5.1 The SD Model Application 238 7.5.2 The GP Model Application 248
7.6 Concluding Remarks 264
CHAPTER 8 FUZZY GOAL PROGRAMMING (FGP) MODEL
8.1 Introduction 265
8.2 Need for Fuzzy Goal Programming Approach 265 8.3 Theoretical Background 266
8.4 Formulation of a FGP Model for Portfolio Managemen1976
8.5 Case Study 279
8.5.1 The Model 279
8.5.2 Analysis of Results 287
8.6 Concluding Remarks 289
CHAPTER 9 SYNTHESIS OF THE PROPOSED DECISION SUPPORT SYSTEM
9.1 Introduction 291
9.2 Inputs and Outputs 291
9.3 "What-if- and "What-to-do-if" Capabilities
298 9.4 The Dialogue System Features 302 9.5 Usage and Implementational Aspects 308
9.6 Concluding Remarks 311
CHAPTER 10 CONCLUSIONS
10.1 Introduction 313
10.2 Summary of the Research Work 313 10.3 Significant Research Contributions 324 10.4 Limitations of the study 328 10.5 Suggestions for further work 398
10.6 Concluding Remarks 329
REFERENCES 330
APPENDICES
CURRICULUM VITAE
