International Asset Allocation and Equity Home Bias in Emerging Markets

317 Pages·2013·16.86 MB·English

International Asset Allocation and Equity Home Bias
in Emerging Markets
Thesis Submitted for the degree of Doctor of Philosophy
at the University of Leicester
By
Dalia M.Reda EI-Edel
B.Sc, M.Sc (AinShams University, Egypt)
Department of Economics
University of Leicester
May, 2010
UMI Number: U560491
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
Dissertation Publishing
UMI U560491
Published by ProQuest LLC 2013. Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code.
ProQuest LLC
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106-1346
International Asset Allocation and Equity Home Bias
in Emerging Markets
Abstract
Dalia M.Reda EI-Edel
This thesis investigates equity home bias from an asset allocation perspective in
emerging markets. Firstly, a review of equity home bias in modem finance literature is
presented, followed by a discussion of the relative strengths and weaknesses of
international asset pricing and optimal allocation models. Secondly, the thesis tests static
and conditional Capital Asset Pricing Models (CAPMs) for 23 emerging markets over the
period February 1997 - December 2007. The study reveals little support for the static
CAPM compared to the conditional version; in which the conditional CAPM seems to
explain excess returns’ dynamics and implies higher volatility persistence in emerging
markets compared to developed markets as documented in the literature. Accordingly, the
study employs a modified trivariate generalised autoregressive conditional
heteroscedasticity (GARCH) model for the period April 1994 - July 2008, in order to
estimate time-varying optimal weights in a portfolio of three assets; namely the return on
the domestic index, the return on the US index, and the return on the UK index. The
number of assets in the portfolio is increased to reach 13 assets in some markets through
the estimation of the Dynamic Conditional Correlation (DCC) model denominated in local
currencies and in US Dollars. The three models show that the optimal weights on domestic
equities divert substantially from the actual equity holdings as documented in survey
reports; in addition to the effect of including more assets in the portfolio, and the influence
of exchange rate risk on optimal weights. Lastly, the thesis examines the variables that
influence equity domestic holdings through the panel estimation of the feasible Generalised
Least Squares (GLS) method in order to control for heteroscedasticity. The study suggests
that factors related to information asymmetries, economic risks at home, exchange rate
volatility, and markets’ inefficiencies are the main factors affecting equity domestic bias in
emerging markets.
1
Acknowledgments
I am utterly grateful to Allah for all the blessings I have had through my life. All
appreciation is due to my supervisor Professor Stephen Hall. I cannot thank him enough for
his generosity in the time, help and advice he gave me. It would have never been any close
to possible to complete this thesis without his continuous support and supervision. He has
been and always will be a constant source of inspiration and a great role model. I am truly
indebted to Professor Hall.
I would like to thank also my second supervisor Dr.Carlos Carrillo-Tudela for his
help with MATLAB software, and valuable comments. I also need to thank Dr. Ali al-
Nowaihi for his help in exploring the thesis topic and his academic guidance.
I am enormously grateful to my mother, father and brother Sherief for their
continuous encouragement, advice, motivation and support over the PhD programme
period. This thesis is entirely dedicated to them.
I have been truly blessed and lucky with my friends in Leicester and back home;
their constant encouragement, help and inspiration offered me support and hope to finish
my programme. In first name alphabetical order, I owe a massive thank-you to Ademola
Adeoye, Ahmed Bassouni, Ahmed Farag, Huda Mergerhi, Heba Saleh, Laura Grigolon,
Maisa Almatrafi, Mohamed Kamal, Nada Magdy, Omar Daoud, Rasha Nasser, Rebecca
Hahn, Sonia Del Cruz, Steven Shimozaki and last but not least Tamer Zaytoun.
I was extremely privileged to meet my colleagues at the University of Leicester.
Their academic and non academic support is indeed immeasurable. I am thankful to
Jonathan Chipili for his advice and feedback on my work; also Marco Grottadaurea for his
enormous helps with software programming. I am deeply grateful to Andrea Oterova, Sahar
Qaqeesh, Francesca Acacia, and Kavita Sirichand for being a family to me. I would like to
thank Dr. Mohamed Shabaan and Sophie N’Jai for their feedbacks on my work, and their
valuable comments.
In addition to all the foregoing, I am very grateful to my sponsor ‘Institute of
National Planning’ in Cairo for the financial support through the Egyptian government. I
also feel deeply grateful to Professor Ali Nassar, Professor Saad Hafez and Salwa Hussein
for their academic and personal advices throughout the PhD programme.
2
List of contents
Abstract........................................................................................................................................................1
Acknowledgments......................................................................................................................................2
Chapter One: Introduction.............................................. 9
1.1 Background of the study............................................................................................................................................9
1.2 Motivation for the study..........................................................................................................................................13
1.3 Objectives of the study............................................................................................................................................17
1.4 Main contributions of the thesis...........................................................................................................................18
1.5 Structure of the thesis.............................................................................................................................................20
Chapter Two: Review of Literature............................. 21
2.1 International Asset Pricing Models: Theory and Evidence in Emerging Markets................................21
2.1.1 Unconditional and Multifactor CAPM.....................................................................................................21
2.1.2 Conditional CAPM..........................................................................................................................................29
2.2 The use of time-varying GARCH models inp ortfolio asset allocation.....................................................37
2.2.1 Multivariate GARCH.......................................................................................................................................38
2.2.2 The use of conditional correlation models in optimal portfolio allocation.............................43
2.3 Equity home bias.......................................................................................................................................................48
2.3.1 Theoretical background...............................................................................................................................48
2.3.1.1 The mean-variance approach.....................................................................................................................48
2.3.1.2 Consumption-based approach.................................................................................................................52
2.3.2 Empirical Studies.............................................................................................................................................55
Chapter Three: Testing the Conditional Capital Asset
Pricing Model in Emerging Markets................................ 59
3.1 Introduction.................................................................................................................................................................59
3.2 Main characteristics of emerging stock market returns.............................................................................60
3.2.1 Definition of emerging stock market......................................................................................................60
3.2.2 Importance of emerging markets............................................................................................................61
3.2.3. Stock indexes benchmarks..........................................................................................................................62
3.2.4. Distribution of returns and CAPM............................................................................................................63
3.2.4.1 Mean returns on emerging markets......................................................................................................66
3.2.4.2 Volatility in emerging markets.................................................................................................................68
3.2.4.3 Normality of returns.....................................................................................................................................70
3.2.4.4 Predictability of returns...............................................................................................................................72
3.2.4.5 Unconditional correlation between emerging equity returns and the MSCI World return
..............................................................................................................................................................................................74
3.3 Testing the unconditional CAPM in Emerging Countries............................................................................78
3.3.1 The Dataset.......................................................................................................................................................78
3.3.2 Estimation of unconditional CAPM in emerging markets................................................................78
3.4 Conditional CAPM in Emerging Countries........................................................................................................81
3.4.1 Estimation of conditional CAPM in emerging markets....................................................................82
3
3.4.1.1 Results for Latin America............................................................................................................................87
3.4.1.2 Results for South and East Asia...............................................................................................................88
3.4.1.3 Results for Emerging Europe......................................................................................................................90
3.4.1.4 Results for Middle East and Africa..........................................................................................................91
3.4.2 Diagnosis Test.................................................................................................................................................93
3.5 Conclusion....................................................................................................................................................................94
Chapter Four: Time-varying optimal weights for
international asset allocation in Emerging Markets.....114
4.1 Introduction................................................................................................................................................................114
4.2 The use of the Multivariate GARCH Model in international asset allocation....................................115
4.2.1 Econometric Model.....................................................................................................................................117
4.2.2 Empirical results...........................................................................................................................................118
4.2.2.1 Results for Latin America markets........................................................................................................118
4.2.2.2 Results for Middle East and African market......................................................................................120
4.2.2.3 Results for Emerging European markets:............................................................................................121
4.2.2.4 Results for East and South Asia markets.............................................................................................123
4.2.3 Diagnostic test for the trivariate GARCH model..............................................................................124
4.2.4 Time-Varying optimal weights according to the trivariate GARCH model.............................125
4.2.4.1 Results for Latin America..........................................................................................................................127
4.2.4.2 Results for Middle East and Africa:.......................................................................................................128
4.2.4.3 Results of Emerging European Markets..............................................................................................128
4.2.4.4 Results of East and South Asia................................................................................................................129
4.2.5 Sharpe Ratio of optimal portfolios constructed with the trivariate GARCH estimation
denominated in local currency.....................................................................................................................................130
4.2.6 Portfolio International diversification from the International Investment Position..............132
4.3 The use of Dynamic conditional correlation model...................................................................................133
4.3.1 Econometric Model.....................................................................................................................................134
4.3.2 Data...................................................................................................................................................................137
4.3.3 Empirical results of DCC estimation denominated in local currency.........................................140
4.3.3.1 Results for Latin America..........................................................................................................................140
4.3.3.2 Results for Africa.........................................................................................................................................141
4.3.3.3 Results for Emerging Europe..................................................................................................................141
4.3.3.4 Results for South and East Asia..............................................................................................................142
4.3.4 Diagnostic test for the DCC model denominated in local currency..................................................142
4.3.5 Time-Varying optimal weights based to the DCC model denominated in local currency 144
4.3.5.1 Results for Latin America..........................................................................................................................144
4.3.5.2 Results for Africa.........................................................................................................................................145
4.3.5.3 Results for Emerging Europe...................................................................................................................146
4.3.5.4 Results for South and East Asia.............................................................................................................148
4.3.6 Sharpe Ratio of optimal portfolios constructed with the DCC estimation denominated in local
currency.................................................................................................................................................................................150
4.3.7 Empirical results of DCC estimation denominated in US Dollar.........................................................151
4.3.7.1 Results for Latin America..........................................................................................................................151
4.3.7.2 Results for Africa.........................................................................................................................................152
4
4.3.7.3 Results for Emerging Europe..................................................................................................................153
4.3.7.4 Results for South and East Asia.............................................................................................................154
4.3.8 Diagnostic test for the DCC model denominated in US Dollar...........................................................155
4.3.9 Time-Varying optimal weights according to the DCC model denominated inU S Dollar...........155
4.3.9.1 Results for Latin America..........................................................................................................................155
4.3.9.2 Results for Africa.........................................................................................................................................157
4.3.9.3 Results for Emerging Europe..................................................................................................................158
4.3.9.4 Results for South and East Asia.............................................................................................................159
4.3.10 Sharpe Ratio of optimal portfolios constructed with the DCC estimation denominated in US
dollar.......................................................................................................................................................................................161
4.4 Conditional correlation..........................................................................................................................................161
4.5 Conclusion..................................................................................................................................................................165
Chapter Five: Equity home bias in Emerging markets:
Panel Estimation................................................................. 239
5.1 Introduction................................................................................................................................................................239
5.2 Main Empirical studies on equity home bias...................................................................................................240
5.2.1 Measuring equity home bias....................................................................................................................240
5.2.2 Explaining equity home bias....................................................................................................................241
5.2.2.1 Hedging against home-country risks...................................................................................................241
5.2.2.2 International diversification gains and costs.....................................................................................245
5.2.2.3 Degree of uncertainty in measuring home bias...............................................................................247
5.3 Econometric Model..................................................................................................................................................250
5.3.1 Controls on capital flows...........................................................................................................................252
5.3.2 Information Asymmetries.........................................................................................................................254
5.3.3 Economic and institutional risks............................................................................................................255
5.3.4 Financial development...............................................................................................................................257
5.3.5 Others..............................................................................................................................................................258
5.4 Model specification.................................................................................................................................................260
5.4.1 Test for cross-panel heteroscedasticity................................................................................................263
5.4.2 Test for Serial Correlation...........................................................................................................................266
5.4.3 Test for Multicollinearity............................................................................................................................267
5.4.4. Test for Omitted Variables.........................................................................................................................269
5.5 Conclusion..................................................................................................................................................................269
Chapter Six: Conclusion.............................................. 292
6.1 Introduction ..............................................................................................................................................................292
6.2 Summary of Findings and contributions.......................................................................................................292
6.3 Limitations of the study and Suggestions for further work......................................................................298
References.......................................................................299
5
List of Tables
Table 1.1: Actual shares of domestic equity holdings to total holdings in emerging markets during
2001-2007.................................................................................................................................................................14
Table 3.1: Characteristics of emerging stock markets returns denominated in US Dollar...................96
Table 3.2: Unconditional correlation coefficients between monthly country index returns for the
period Jan 1997-Dec 2007...................................................................................................................................99
Table 3.3: Unit root tests of excess returns on countries indexes, and the market premium 101
Table 3.4: Estimation results for BEKK bivariate GARCH [the number in parentheses denotes p-
value].......................................................................................................................................................................102
Table 3.5: The Ljung-Box Statistics of Scaled Residuals.........................................................................105
Table 4.1: Unit root tests of returns on markets indexes used for the trivariate GARCH..................168
Table 4.2: Estimation of the trivariate GARCH. A =h +rt 1 + A i +&,- ~*~£, t Tlie number in
3-
parenthesis is the p-value].................................................................................................................................169
Table 4.3: The Ljung-Box statistics of squared scaled residuals of domestic returns.........................173
Table 4.4: Mean values for optimal weights according to the trivariate GARCH estimation 176
Table 4.5: The share of US and UK equities in the international investment position of emerging
markets to domestic market capitalisation (In Percentage)........................................................................177
Table 4.6: Unit root tests of returns on markets indexes for the use of DCC estimation...................179
Table 4.7: DCC Model estimation denominated in local currencies......................................................181
Table 4.8: Ljung-box Q-statistic for the standardised residuals For DCC estimations denominated in
local currency........................................................................................................................................................184
Table 4.9: Mean values for time-varying optimal weights for DCC model denominated in local
currencies:.............................................................................................................................................................194
Table 4.10: Ljung-box Q-statistic for the 1st and 12th orders in levels of the standardised residuals
according to the DCC estimation denominated in US Dollar.....................................................................197
Table 4.11: Mean values for time-varying optimal weights for DCC estimation in US Dollars:... 199
Table 5.1: Data Description and Sources.......................................................................................................271
Table 5.2: Summary Statistics for variables used during the period 2001-2007................................ 274
6
Table 5.3: Optimal weights of domestic equities (Using DCC model estimation in local Currency)
.................................................................................................................................................................................275
Table 5.4: Actual holdings of domestic equities to total equity holdings................................................275
Table 5.5: The extent of domestic equity bias (The difference between the actual share of domestic
equity holdings and optimal domestic equity weights)..............................................................................276
Table 5.6: Construction of capital outflow control index (CONT).........................................................277
Table 5.7: Panel Estimations (Dependent variable: Ratio of domestic equity holdings to total equity
holdings).................................................................................................................................................................291
7
List of Figures
Figure 3.1: Market Capitalisation in Emerging regions through the years 2004-2007..................... 107
Figure 3.2: Emerging stock markets returns denominated in US Dollar (Feb 1997- Dec 2007).... 108
Figure 3.3: Time-Varying Beta based on the bivariate GARCH estimations.......................................110
Figure 3.4: Time-varying Covariances/variances based on the bivariate GARCH estimations.......112
Figure 4.1: Expected Returns and Conditional Variances according to the trivariate GARCH
estimation..............................................................................................................................................................202
Figure 4.2: Time-varying weights in the optimal portfolio........................................................................210
Figure 4.3: Sharpe ratios for optimal portfolios versus domestic based on trivariate GARCH
estimations.............................................................................................................................................................214
Figure 4.4: Time-varying optimal weights according to DCC model estimation denominated in local
currencies...............................................................................................................................................................217
Figure 4.5: Expected returns and conditional variances on selected cases according to the DCC
estimations.............................................................................................................................................................222
Figure 4.6: Sharpe ratios for optimal portfolios versus domestic equities resulted from the DCC
model dominated in local currency.................................................................................................................230
Figure 4.7: Time-varying optimal weights according to DCC model estimation denominated in US
Dollars....................................................................................................................................................................232
Figure 4.8: Sharpe ratios for optimal portfolios versus domestic equities resulted from the DCC
model dominated in US Dollars........................................................................................................................236
Figure 4.9: Conditional Correlations for selected markets resulted from the DCC model in US
Dollars:...................................................................................................................................................................238
8