Banking industry volatility and economic growth

Particularly, we pay more attention on the way country characteristics, such as the effect of low and high inflation, Worldwide Governance Indicators WGI from the updated database of Kau

UNIVERSITY OF ECONOMICS ERAMUS UNIVERSITY ROTTERDAM

HO CHI MINH CITY INSTITUTE OF SOCIAL STUDIES

VIETNAM – NETHERLANDS

PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

BANKING INDUSTRY VOLATILITY

AND ECONOMIC GROWTH

A thesis submitted in Partial Fulfillment of the Requirements for the Degree of

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

By

TRAN QUOC THANH

Academic Supervisor:

Assoc Prof Dr VO XUAN VINH

HO CHI MINH CITY, December 2015

30th November 2015

Trần Quốc Thanh

to all my questions and queries punctually I could not have imagined having better supervisors and advisors for my research

Besides my mentors, I would like to thank Dr Pham Khánh Nam, Dr Dương Như Hùng, and Prof Dr Ardeshir Sepehri for their thorough comments and worthy ideas that help to enhance my thesis’s value

My sincere thanks also goes to all the lecturers at the Vietnam – Netherlands Program for their knowledge of all the courses, during the time I studied

at the program I would like to offer my special thanks to lecturers in Data Center in University of Economic and Law, Dr Lê Văn Chơn, Dr Trương Đăng Thụy, lecturer Hoàng Trọng, who help me significantly in the courses and thesis writing processes In addition, I would like to express my great appreciation to

my friends for their motivations

Last but not the least; I owe a very important debt to my family for giving birth

to me at the first place and supporting me spiritually throughout my life

Hồ Chí Minh city, December 2015

ABSTRACT

There is growing evidence from multi- studies indicating that there are lots of determinants advocate to economic growth However, very few research papers contribute to banking sector, vital field of modern economy It is unclear whether it

is appropriate to assume an identical turning point in the banking industry volatility and growth relation divided into across income criteria and geographical region criteria In this research, we keep investigating the relationship between banking volatility and economic growth in detail ways after examining carefully the studies

of Moshirian & Wu, (2012); Lin & Huang, (2012) Using GMM techniques for dynamic panel data to analyze one main group and five subsamples: all 22 economies, 11 upper middle income economies, 11 low income and lower middle income economies, 8 Sub-Saharan Africa economies, 6 South Asia and East Asia economies, 5 Latin America economies, by using dynamic panel techniques to analyze panel data

Particularly, we pay more attention on the way country characteristics, such as the effect of low and high inflation, Worldwide Governance Indicators (WGI) from the updated database of Kaufmann (2013) and financial development characteristics influence the relationship between bank volatility and economic growth The quarterly panel dataset, which is available and easy approach from international Datastream

The simple correlation between GDP growth rates and banking volatility is slightly higher in geographic region groups There is relationship of banking industry volatility and economic growth in all 22 economies, and in five subsamples divided into income criteria and geographical region criteria, even in the presence of market excess returns, and the relationship between banking volatility and economic growth is affected by the country characteristics and financial development when the interaction terms have statistical significant Except for Voice and Accountability having no effect Some research papers of Fama (1981, 1990) and Schwert (1990) have proved that the effect of the uncertainty of banking industry on economic growth is uncorrelated with the effect of the market stock return in

general on economic growth Hence, our results is more one evidence for the relationship between the stock returns of bank and economic growth

Key words: banking volatility, difference GMM, system GMM, country

characteristics, financial development characteristics, effect of inflation

TABLE OF CONTENTS

CHAPTER 1: INTRODUCTION

1.1 Problems statement: 6

1.2 Research objectives: 8

1.3 Main research questions: 8

1.4 Structure of the thesis: 9

CHAPTER 2: LITERATURE REVIEW 10

2.1 Banking industry volatility and economic growth: 10

2.2 Indicators of country characteristics: 12

2.3 Financial indicators and real activity: 16

2.4 Stock markets and economic growth: 17

2.5 Conceptual framework: 19

CHAPTER 3: METHODOLOGY, MODEL SPECIFICATION AND DATA 20

3.1 Data: 20

3.2 Methodology: 27

CHAPTER 4: RESULTS AND FINDINGS: 30

4.1 Descriptive statistics of variables: 30

4.2 Econometric results: 33

CHAPTER 5: POLICY IMPLICATION, CONCLUSION AND LIMITATION: 60

4.1 Policy implication and conclusion: 60

4.2 Limitation of the research: 62

REFERENCES: 64

APPENDICES :

Table of empirical studies relating to economic growth rate: 68

Appendix 1: Full sample of all 22 economies: 74

Appendix 2: 11 Upper middle income economies: 98

Appendix 3: 11 Low income and 11 lower middle income economies: 122

Appendix 4: Africa economies: 146

Appendix 5: Asia economies: 170

Appendix 6: Latin America economies: 194

ABBREVIATIONS

WGI: World Governnance Indicator

EMH: The Efficient market hypothesis

GDP: Gross Domestic Product

GNI: Gross National Income

GMM: The Generalized Method of Moments Estimation

GMM(DIF): The Difference Generalized Method of Moments Estimation

GMM(SYS): The system Generalized Method of Moments Estimation

CHAPTER 1: INTRODUCTION

1.1 Problems statement

Very few research papers contribute to banking sector, vital field of modern economy It is evident that numerous research papers for many decades have shown financial development as important channel for economic growth Banks play a crucial role in the economic growth of a country by allocating funds among all sectors, primary sectors, secondary sectors, tertiary sector, etc Most of existing researches focus on performance of the bank, liquid liabilities to GDP, market capitalization per GDP, credit to private sector per GDP, etc as factors of financial development Bank sector play crucial role in supply facilities through deposit and lending, credit, banking services, money transfer, etc to economic activities However, very few researches measure directly the effect of banking operation on economic growth Serwa (2010) indicates that banking crises cause output growth

to slow down A well-functioning banking system facilitates infrastructure for other sectors running smoothly Therefore, banking stock return will be reflected in the quality of bank credits According to Bruner & Simms (1987); Cornell & Shapiro

1986, the market for commercial bank securities operating are efficient and contain information about the quality of bank loan portfolios

There are close relationship between bank stock returns and economic growth Base

on asset-pricing theory, and on many researches of economists Cole, et al (2008); Moshirian & Wu, (2012); Lin & Huang, (2012), prove that stock returns of banking industry reflecting the performance of the bank can predict economic growth In addition, according to the view of market efficiency, at any point in time, prices of securities in efficient markets reflect all known information available to investors

In other words, the expected future cash flows of the banks are reflected in the present stock price This depends on efficiency of loan projects Bank stock returns will reflect the efficiency of the market in using funds to investment Furthermore,

in most of countries, commercial banks, PLCs, are broadly representative of country’s banking sector since they account for very high position in the whole banking system Consequently, there are correlation between bank stock returns and future economic growth

In some research of Cornett (2010) and Naceur and Ghazouani (2007), institutional framework, such as country specific, financial system indicators also have significant influence on banking operations Moreover, in the investigation of Asante, S., Agyapong, D., & Adam, A M (2011), country characteristics help banks operate smoothly as well as improve their services This promotes economic growth significantly In the indicators representative country specific, the negative effects of inflation have been studied in a lot of models of economic growth, it undermines the confidence of domestic and foreign investors as well as consumers about the future economic growth (Andrés & Hernando, 1999) Secondly, the sustainable increase in living standard for a country means a larger voice on the world stage There are a lot of measures of the quality of governance have been built to evaluate of the quality of governance, among these are the Worldwide Governance Indicators, six institutional variables rank countries on six aspects of good governance (Kaufmann, 2013) Besides, the impact of banking stock returns

on economic growth is captured by country characteristics and financial development (cole, et al, 2008; Moshirian & Wu, 2012)

According to the point of view that banking operation contains information about performances of a lot of sectors reflecting the health of the economy (cole, et al, 2008; Moshirian & Wu, 2012; Lin & Huang, 2012) It is indicate that the relationship between banking industry volatility and economic growth that is independent of the information contained by overall market returns Since the volatility of the bank relate to the variation of stock returns of the banking industry which refer to each individual bank Therefore, this information should be independent of information being reflected in market excess returns which is representative for the whole public limited company (PLCs) in the stock market (Oshiriana & 2012; Lin & Huang 2012) Similarly, Naceur & Ghazouani (2007) indicate that the impact of equity market on growth is independent to the impact of bank development on growth Publicly traded banks also account for high proportion in the whole, this lead to banking industry stock returns will represent for the whole banking sector in the most of countries

In the researches on banking industry volatility and economic growth topic Moshirian & Wu, (2012), the investigators concentrate more on country classified

criteria by compare two subsamples, developed markets and emerging markets, which are nations with social or business activity in the process of rapid growth and industrialization, and Pei-Chien Lin, Ho-Chuan Huang, 2012 pay more attention on the sample in developed countries and middle income countries Whereas, samples surveyed in my study have different approach, collected data is based on income criteria and geographical region criteria, and data investigated is divided into one main sample including all 22 markets and five subsamples including11 upper middle income group, 11 low income and lower middle income group, 8 Sub-Saharan Africa group, 6 South Asia and East Asia group, 5 Latin America

The extreme volatility of banking industry stocks has trigged for the confusion in finance performance and the economic crisis after that All the components of volatility including firm volatility, industry volatility, marketing volatility are countercyclical and tend to lead variation in GDP (Campell, 2001) In this study, we keep investigating the uncertainty of bank stock price in the relation with the markets behaving erratically in the financial markets as well as in economic growth

We also survey the effect of low and high inflation on economic growth in the interaction with banking volatility Thirdly, we examine the effect of country characteristics, which is Worldwide Governance Indicators, and financial development variables impact on the relationship between bank volatility and future economic growth rate By doing this research, we hope to fill the gap of present researches, and contribute some small findings to these economies

1 2 Research objectives:

- To investigate the relationship between banking industry volatility and economic growth in full sample and five subsample divided basing on income criteria and geographical region criteria

- To examine whether country characteristics and financial development impact on the relationship between bank volatility and economic growth

1.3 Main research question:

- Is there any relationship between banking industry volatility and economic growth

in selected samples?

- Whether country characteristics and financial development characteristics strengthen or weaken the relationship between bank volatility and economic growth

in selected samples?

1.4 Structure of the thesis:

After the finish of Chapter 1, the rest of this thesis will be categorized as following chapter:

Chapter 2 introduce banking industry volatility and economic growth, financial indicators and real activity, indicators of country characteristics, stock markets and economic growth, relevant literature reviews of these relationships

Chapter 3 states methodology, model specification and data scope used This chapter also gives readers clear explanatory variables used, suggested statistical diagnostic of significance of variables Simultaneously, data scope and sources together with model conceptual framework

Chapter 4 interprets results and findings of thesis regression model

Chapter 5 concludes with thesis limitation and further research suggestion

CHAPTER 2: LITERATURE REVIEW

This chapter shows out some appropriate definitions of banking industry volatility, indicators of country characteristics, such as Voice and Accountability, Free election, Political Stability and Absence of Violence, Government Effectiveness, Regulatory Quality, Rule of Law, high inflation, low inflation, financial development from the perspective of understandability and data availability Simultaneously, this section also demonstrates and discusses in terms of approaching datasets collected, researcher, methods applied Therefore, this chapter includes five parts.

2.1 Banking industry volatility and economic growth

Base on asset-pricing theory, firstly, standing for the intergration with the world stock market, refers to the price risk as well as development of the stock market Deveneux and Smith (1994), and Obstfeld (1994) state that intergrated stock market should increase the risk diversification, thus increase economic growth Secondly, overvalue price relates the price of any asset to its future dividend, and thus incorporate way to estimate the time and risk in asset valuation, prove that stock returns of banking industry reflecting the performance of the bank can predict future economic growth

In addition, according to The Efficient market hypothesis (EMH), the view of market efficiency, share prices always being traded at their fair value incorporate and reflect all relevant information in the market At any point in time, prices of securities in efficient markets reflect all known information available to investors

To be more clearly, the expected future cash flows of the banks are reflected in the present stock price This depends on efficiency of loan projects Bank stock returns will reflect the efficiency of the market in using funds to investment Furthermore,

in most of countries, commercial banks, PLCs, are broadly representative of country’s banking sector since they account for very high position in the whole banking system Consequently, there are correlation between bank stock returns and future economic growth

The global financial crisis launched firstly in the US spread to financial markets around the world with banking crisis in 2008 This leads to weak operation in supply commodities to out-banking sectors and harm the aggregate output Laeven

& Valencia (2010) find that reducing in output happening after banking crisis average around 37% of potential output Campello et al., (2010) survey 1,050 Chief Financial Officers in American They indicate that firms do not have ability to borrow externally during the credit crisis of 2008 This leads them to fail to catch investment opportunities, with 86% of constrained U.S In short, the weakened function of banking system cause negative effect on the flow of economic activities This financial shock may cause very high inflation rate in economies surveyed after that, and high interest rate simultaneously

In the research on Banking industry volatility and growth of Pei-Chien Lin, Chuan (River) Huang, 2012 They use Industry, Country

Ho-External Finance, Bank Development, Volatility of Bank Development, Stock Market Development, Volatility of Stock Market Development as independent variables by difference-in-difference framework of RajanandZingales (1998) to the cross-country, this study cover 19 years from 1980 to 1999 for developed countries and middle income countries They shows that banking-industry-volatility, estimated as the standard de viation of the growth of privatecredit, effect on the growth of industries negatively that are more externally financially dependent However, the detrimentalgrowth effect of banking sector volatility disappears when thesample is restricted to the relatively placid 1980.Compared to the 1980s, the 1990sare characterized by a moreeconomically integrated world accompanied by moreoften unpredicted financial crises that disturb the banking sector As such, this results imply that in a moreeconomically integrated world, the stabilityof bank development may be important to long-run growth On the other hand, banking industry instability relate to market performance where contain the economy figures The uncertainty tends to decrease as the stock market rises or output grows

up and incline as the stock market depress or output slump (Crestmont, 2011) For the research of Fariborz Moshiriana and Qiongbing Wu, 2012 when they use OLS difference GMM and Systerm GMM technique to survey in one full sample of 36 developed countries, and two subsamples of 18 developed economies, 18 emerging

economies in the period of 33 years from 1973 to 2006 These investigators found that banking industry volatility reflects on lots of information about the health of future economy It represent for thestability of the performance ofthe bank Volatility works well to help to identify the bank’ performance bottoms base on high volatility While the country-specific and banking institutional characteristics affect these relationships They also prove that there is negative relationship between bank volatility and future economic growth Banking crisis and banking accounting standards mak e the relationship between bank-excess- returns and future economic growth to be positive an d make the relationship of bank- volatility and future economic growth negative more tremendously However, government ownership reduce these relation In contrast, the enforcement of the insider- trading law increase the first relation positively and decline the second

relation negatively The three indicators of financial d evelopment strengthen the firstrelation and impact ambiguously on the second

2.2 Indicators of country characteristics

In the research of Cornett, 2010, institutional framework also has significant impact

on banking system To be more detail, political bureaucrats weaken the performance

of state-owned banks, and the state-owned banks normally run less profitably, held less core capital, and contained greater credit risk than private-owned banks Moreover, Naceur and Ghazouani (2007), who use Panel data with GMM estimations when survey 11 MENA countries prove that the impact of equity- market on growth is independent to the impact of bank development on growth, and Financialdevelopment keep unimportant role for economic growth This weak relationship is harmful for the economy Besides, there is negative relationship between bank-development and economic growth in conditionofcontrolling for the stock market development since unwell-functioning financial system and poor institutional environment Therefore, institutional environment, such as country specific, financial system indicators have important contributions to institutional framework as well as the banking operation They make the relationship between

bank development and economic growth negative in condition of controlling for the stock market development since unwell-functioning financial system as well as poor institutional environment Besides, in some research (Perez, et al., 2015; Filippidis,

et al., 2014; Cole, et al., 2008 ) tell us the same result that un-well financial environment harms the banking sector development This depends on specific structure as well as institutional framework in each country Furthermore, in the investigate of Asante, S., Agyapong, D., & Adam, A M (2011) when they use market capitalization, bank competition, Domestic bank credit as percent of GDP as independent variables to cover the study in 17 years from 1992 to 2009 in Ghana by Autoregressive Distributed Lag (ARDL) and Dynamic OLS (DOLS) method This research prove that banking institutional characteristic, banking competition and stock market development make smoothing operation, improving services in the banks This affects economic growth positively as a results

The macroeconomics theory has long discussed the causality between inflation and economic growth for over half century To be more clearly, Philips model and Keynessian’ view stated that there are strong relationship between economic growth and inflation rate in every country They have positive link in the case of high inflation shift the aggregate demand to the right Conversely, if high rate in inflation shift aggregate demand to the left, this relation is negative These circumstance present in the case of monopolistic pricing, wide volatility in exchange rate, or the shocks in input supply This mean that the economic will suffer difficulty Especially, it may get high unemployment rate when inflation rate is higher than 2-3 percent According to Bruno, M., & Easterly, W (1998), the separate the high grade

in the inflation rate (the grade at which the inflation rate exceed calculated threshold

is estimated 40 percent yearly) weaken the previous research results and give out one new exploration That is the high grade in inflation harm the economic growth, and otherwise, the low grade in inflation boost the economic in one country These variables have negative relationship, especially in the long stage is supported by Andrés & Hernando, (1999)

The sustainable increase in living standard for a country means a larger voice on the world stage There are a lot of measures of the quality of governance have been built to evaluate of the quality of governance, among these are the Worldwide Governance Indicators, six institutional variables rank countries on six aspects of good governance (Kaufmann 2013) The effect of governance quality on economic performance is supported by Kong, T (2011) Miyazawa, I., & Zusman, E (2015) prove that government effectiveness and rule of law have substantively and statistically significant effects on progress countries, control of corruption has positive correlations with the government effectiveness However, voice and accountability” and “control of corruption” do not have the signif icant effects on activities of one country In contrast, Kaufmann, D., Kraay, A., & Mastruzzi, M (2010) argue that construct validity is not a useful tool to assess the merits of the WGI In other research paper of Omoteso, K., & Ishola Mobolaji, H (2014) Author evidence out difference results, political stability and regulatory variables impact on economic growth in the separate region (Sub-saharan Afica) positively and significantly, this is more strongly for the variables of voice and accountability and rule of law, but government effectiveness have negative impact

Voice_and_Accountability: “The variable measure the degree to which their citizens may present in election for authorities in one country, freedom of voice, free media” Free_media, freedom of expression mitigate the asymmetric information problem In controlled media countries, rumors have considerable influence on the financial-market Thus free_media partially weaken the effect of banking industry volatility on economic growth Free_election, freedom of association associated with free media may impose a balance and check on political power and prevent government officials, interest groups from implementing distortionary macroeconomic policies that make the country more vulnerable Free_election, free media help constrain politicians Thus free election weaken the effect of banking industry volatility on economic growth

Political_Stability and Absence_of_Violence: “The variable measure perceptions of the likelihood that the government will be destabilized by unconstitutional or violentmeans, including politically_motivated violence and terrorism” Political_stability and absence_of_violence may lead to less volatility in policies

toward businesses (banking and non-banking) in the future, less volatility in the economy It is less likely to lead to capital flight or panic selling of currency Thus Political Stability and Absence of Violence weaken the effect of banking industry volatility on economic growth

Government_Effectiveness: “The variable measure the public quality in serving its citizens, and the extent of its independence, the policy quality, and the commitment

of authorities to make their policy occur in the real life The more effectiveness, the less vulnerability of financial field” A higher level of quality of public services, higher degree of its dependence from political pressure, there will be less volatility

in the government conduct of economic policies/activities The financial market may be less vulnerable to panic and herding Thus, Government Effectiveness weaken the effect of banking industry volatility on economic growth

Regulatory_Quality: “The variable measure the ability of authorities to make the regulations feasible and improve the private field” In emerging market economies, the operation of state owned firms is normally inefficiency This affects outcome of the economy negatively Sound policies enhancing private sector development would boost economic performance of a developing economy, improve confidence

in the financial market Thus Regulatory Quality weaken the effect of banking industry volatility on economic growth

Rule_of_ Law: “The variable measure the degree to which government make the quality of policy, courts, crime laws, violence, enforcement of contract, property rights feasible” Higher confidence level in the law system is formed from the higher grade of Rule_of_Law, effective contract enforcement, fair and impartial court, protection of property right Higher grade of this variable strengthens trust in the economy This contribute insignificantly to volatility in financial market Thus Rule of Law weaken the effect of banking industry volatility on economic growth

Control_of_Corruption: “The variable measure the degree to which authorities carry out power to public” Economic activities, such as import/export activities, manufacturing activities, government activities to banking activities, are impacted seriously by corruption, from that may decline the efficiency of economic performance High level of corruption may cause inefficiencies and misallocation of

resources running up to poor economic outcome Economy get vulnerable to capital flight easily because of getting stuck in economic performance Thus, Control of Corruption weaken the effect of banking industry volatility on economic growth

2.3 Financial indicators and real activity:

In the four theoretical views, which highlight the impact of financial development

on economic growth, the bank based theory, the market based theory, the finance and services theory and law-finance theory They point out that financial development happening in the case of increase ịn quality, quantity and efficiency of financial system This performance involves the connection of many activities and institutions Financial instruments, markets and intermediaries perform more efficiency make financial system develop, then they affect the information, enforcement and transaction costs (Levine, 2005) Therefore, financial development comprise the enhancement in allocating resource, monitoring investment, mobilizing saving, diversifying and managing risk Furthermore, financial system is enhanced thanks to the improvement in controlling corporate governance after providing finance and the facilitation in exchanging goods and services Each of these performance is likely to influence saving and investment decisions, thereby they affects economic output The importance of financial intermediaries to economic growth appears in the research of King & Levine, (1993a); Levine et al., (2000); McKinnon, (1973); Shaw, (1973) They evident that the key factors for different economic growth of each country derives from the differences in the quality and quantity of services provided by financial intermediations

Three indicators represent for the development of financial system: (1) Domestic- Credit to private field, (2) Liquid-Liabilities (M3) as % of GDP, (3) the ratio of Stock-Market-Capitalization to GDP Firstly, financial resources mostly of corporations are funded to private field in the way of loans, non-equity securities, credits for commercial, accounts receivable These form Domestic- Credit to private field Secondly, the ratio of Liquid-Liabilities in financial system to GDP M0 (the total value of currency and deposit in the central bank) plus M1 (deposits and electronic currency) plus M2 (time and savings deposites and other deposit for transferable foreign currency, certificates as well as securities repurchase

agreements) plus checks for travelers, paper for trades, time deposits for foreign currency, and share of funds for the market It implies that the depth of financial intermediaries is positively related to the provision of financial services; the stock market contributes significantly to the development of financial sectors and economic growth (Pagano, 1993) The appearance of stock market can help investors holding financial portfolio to reduce transaction costs and diversify their risks Since risk diversification incentives investors to hold more their individual assets in productive capital, it is benefit to boost the economic growth For this reason, the last indicator, which is also used to measure the depth of financial development, is the ratio of stock market capitalization to GDP (stock_cap) (Jun, 2012; Levine & Zervos, 1996; Ndlovu, 2013; Wu et al., 2010) It equals the total value of all listed shares in a stock market as a percentage of GDP

2.4 Stock markets and economic growth

For the developing countries, in the research of Wang & Ajit (2013) using unit root test and the cointegration framwork prove that stock market does not affect economic growth positive in China, this result fit with the report of Harris (1997) for developing countries if the stock market is mainly an administratively driven market Osamwonyi et al., (2013), authors find for Ghana and Nigeria, there is no causal relationship between stock market development and economic growth, but they have a bidirectional causal relationship when they use Granger Causality test procedure as developed in Granger This report is the same to Rahimzadeh, Farzad (2012) using data related to the Middle East and North Africa in the period 1990-

2011 In contrast, in the sample of 35 developing economies, Cooray A, (2010) the empirical findings of the study show that policy measures taken to increase the output, liquidity and activity of the stock market will further increase output growth

In the study about the relation among banks, stock markets and economic growth of Kim, D & Lin, S (2013) who demonstrate that the interaction of each of the three variables is in important ways While both contain information about output growth, banking development contribute more for growth in low-income countries and stock market development is more contribution for output growth in high-income or low- inflation countries The authors explore coexistence of a positive effect of banking

development on stock market development and a negative effect of stock market development on banking development In addition, they also find the evidence for the feedback effects of growth on both banking and stock market development Moreover, Rabiul (2010) using Generalized Method of Moments (GMM) to research sample of 80 developing countries spanning from the period of 1973 to

2002 prove that banks and stock markets have positive impacts on economic growth separately and positively, they are important to boost long-run growth in developing countries In addition, author also indicates the relation between finance and growth

is non-linear In other words, financial development is more beneficial for countries developing from a very low stage; however the marginal benefit to rich countries is less than for low income countries due to diminishing marginal returns

2.5 Conceptual framework

Economic Growth

Moshiriana and Wu, 2012

Country characteristics

Financial development

Banking volatility

Liquid Liabilities

Stock Market Capitalization

Political Stability and

Market excess returns

Domestic Credit to Private

Sector

Rule of Law

Control of Curruption

Inflation 1

Lin and Huang, 2012

Oshiriana & 2012; Lin & Huang 2012

(+)

(+) (+)

(Ambiguous)

(Ambiguous)

CHAPTER3: METHODOLOGY, MODEL SPECIFICATION AND DATA

In previous sections, the impact of banking volatility, banking excess return, market excess return, country specifics, financial development have been introduced and discussed with the purpose is that the readers will have overlook on factors impacting on GDP growth rate This section shows out some following issue: (1) data and sample size, (2) methodology employed

3.1 Data

The data sets include information in income criteria and geographical region criteria

by researching in detail in one main sample and five subsamples: all 22 markets, and 11 upper middle income markets, 11 low income and lower middle income markets, 8 Sub-Saharan Africa markets, 6 South Asia and East Asia markets, 5 Latin America markets as subsamples over the period from 2003 to 2014 in quarter data, with the longest time series being 11 years and the shortest 02 years The sample period for each market is shown in column 4 of Table (3) The selected economies’ data is based on the available data on bank_ equity_price, quarterly macroeconomic time series and short_term interest rates Table (1) summary variable calculated and their sources

The primary variable is banking industry volatility There are many way to estimate the volatility of the bank However, in this study, the more correct way is applied in detail That is a disaggregate approach base on the method of Campell et al (2001)

to calculate the banking volatility variable, we carry out in some steps First of all,

we calculate the portfolio of listed banks for each market collecting in International Datastream sources Basing on available Market price data, the maximum number

of 30 listed banks for Indonesia, and the minimum of 2 for Mauritius and Uganda Nevertheless, when collecting available Market Capital data, we have the maximum number of 30 listed banks for Indonesia and the minimum of 1 for Philippines In this case, Philippines has complete data set of market price of 19 banks, quarterly GDP series and short-term interest rates However, it has market capital for 1 listed banks in the Datastream banking sector, so the sample is analyzed on available data

in these indicators of individual banks Since all 22 countries have market economies, we only collect available banks on domestic stock exchange market The banks running in domestic market and in foreign markets, but listed in stock exchange market abroad exclude in our sample Therefore, few banks can representative for the whole market These variable, the interest rates, GDP series and the market price index for each country also are extracted from Datastream The sources of researching data collected is diversity Table (1) summary information about variables calculated in this research With the purpose of prolong the time- series information in this study, we handle yearly data by overlapping method with observations in quarter

Second of all, this paper use the continuous return over Rf (risk-free-rate) to measure the excess-return in weighted value on the portfolios lists of the bank for the every economy which is estimated This research collect Treasury-Bill rate in three month or Deposit-rate in three month depending on the data in Datastream

We use MC (Capitalization) to estimate the weights The Capitalization of bank-j over the total Market-Capitalization of the banking field at the end of the period (t-1) and remain constant within period (t) build the weight for bank-j Third of all, Excess-Return is built on the market index of each economies,

Market-in the next stage, the regressions of Excess-Return of the bank Market-in quarter data against the Excess-Return of the market in quarter data is conducted to get the bêta

of each economy, which is supposed constant over the sample period for each economy, but it is assumed to vary timely in long run Nonetheless, this study analyzes large amount of economies It is more sense to simplify for our assumption and running the same model for different economies in the most consistent way

After all above process, we have complete data sets on 22 economies We divide the panel data into one main sample and five subsamples: all 22 markets, and 11 upper middle income group, 11 low income and lower middle income group, 8 Sub- Saharan Africa group, 6 South Asia and East Asia group, 5 Latin America group as subsamples We follow the threshold levels of GNI per capita are those calculated

by the World Bank in 2012 Countries with less than $1,035 GNI per capita are belong to low-income country group, those with between $1,036 and $4,085 as lower middle income country group, those with between $4,086 and $12,615 as

upper middle income country group GNI per capita in dollar terms is estimated using the World Bank Atlas method in 2012

In the third step, we calculate quarterly bank volatility (VOLit) in monthly data, whereas, Moshirian & Wu (2012) calculate it on weekly data, as follow formula: VOLit=Var(Rit)=βim2 VAR(Rmit)+σ^2it

Where:

VAR(Rmit) = ∑m€t (Rmim - µmit)2

σ^2

it = ∑m€t (Rim - βim Rmim)2

βim is the bêta of the banking industry with proxy to the market in economy i Rmim

is the monthly excess market return in economy i µmit is the moving average monthly excess market return for country i over period t (here t is quarter) This calculate way differ from the method of Moshirian & Wu, (2012); Lin & Huang, (2012), who use the average monthly excess market return

Rimis the monthly excess-return in weighted value on the portfolios lists of the bank for the economy-i This study deduct the monthly rate of risk-free, which is getting

by dividing the yearly interest rate in short run by 12 weeks, as a result we have the excess-return for each month Make reference to the method of Cole et al (2008), most of variables is estimated These are continuous real growth rate (Growth), which make up dependent variable, lagged Excess-Return of the market (Rm), which is controlled variable, the characteristics of each country, which are six governance indicators, low inflation, high inflation, and indicators of financial development, which are domestic credit to private sector, liquid liabilities

The indicators presenting characteristics of each country relate to economic growth

or the efficiency of the economy in one country in long stage These indictors respect the difference in the cross section in institutional framework of one country Next, we estimate the effect of the differences in institutional framework of one country on the relationship between baking industry volatility and economic growth rate

- The banking industry volatility (primary variable) is expected to create negative effect on growth (Pei-Chien Lin, Ho-Chuan Huang, 2012; Fariborz Moshiriana and Qiongbing Wu, 2012)

- The dependent variable is GDP growth rate (Growth) It is calculated by taking logarit of the ratio of GDP at period t and GDP at period t-1 at the constant prices ( Growth=LOG(GDPt / GDPt-1))

- The control variable is lagged market excess return It is defined as the excess return on the market index in country-i It is estimated by taking logarit of the ratio

of market price index at the end of period t and market price index at period t-1 of country i, t is in quarter, then minus the risk-free rate (Rf), which is Treasury-Bill rate in three month or deposit-rate in three month To compromise the empirical results, the impact expected direction ambiguous is indicated by Wang & Ajit (2013), Harris (1997), Osamwonyi et al., (2013), Rahimzadeh, Farzad (2012), Cooray A, (2010) ( Rm=Rmit=log(Pmit/Pmi(t-1)-Rfit ))

Eight country characteristic indicators:

Six Worldwide Governance Indicators (WGI), they are dataset cover some qualify indicators presented the health of Government over the world They range in units from around -2.5 to 2.5, with higher value corresponding to better governance outcomes (Kaufman, 2013) The data is in annual format, so we use the overlapping annual data with quarterly observations to prolong the time-series

- Voice_and_Accountability: This variable do not have the significant effects on activities of one country (Miyazawa, I., & Zusman, E, 2015) But in one of the research of Omoteso, K., & Ishola Mobolaji, H (2014), they evidence that this indicator and economic growth has positive relationship for Sub-saharan Africa region

- Political_Stability and Absence_of_Violence: It is expected to boost the economy and weaken the effect of banking industry volatility on economic growth

- Government_Effectiveness: It has positively and statistically significant effects on progress countries (Miyazawa, I., & Zusman, E 2015), but this is negative in the

research of Omoteso, K., & Ishola Mobolaji, H (2014) for Sub-saharan Africa region

- Regulatory_Quality: It is expected to improve the economy and weaken the effect

of banking industry volatility on economic growth

- Rule_of_Law: It has positively and statistically significant effects on progress countries (Miyazawa, I., & Zusman, E, 2015) It is expected to weaken the effect of banking industry volatility on economic growth

- Control_of_Corruption: It has positive correlations with the government effectiveness (Miyazawa, I., & Zusman, E, 2015)

- Inflation (1) is the dummy variable, it takes on the value of one when the value of inflation is smaller than the sample group (all economies) median and a value of zero otherwise The data is in annual format, so we use the overlapping annual data with quarterly observations The positive sign is expected for the effect

of this variable on growth (Bruno, M., & Easterly, W, 1998; Andrés & Hernando, 1999)

- Inflation (2) is the dummy variable that takes on the value of one if inflation

is greater than the sample group (all economies) median and a value of zero otherwise The data is in annual format, so we use the overlapping annual data with quarterly observations The negative sign is expected for this variables on growth (Bruno, M., & Easterly, W, 1998; Andrés & Hernando, 1999)

Three financial development indicators

- Domestic-Credit to private field is defined as financial resources mostly of corporations, which are funded to private field in the way of loans, non-equity securities, credits for commercial, accounts receivable It is benefit to improve the economic Thus, the positive sign is expected for this indicator

- The ratio of Liquid-Liabilities in financial system to GDP M0 (the total value of currency and deposit in the central bank) plus M1 (deposits and electronic currency) plus M2 (time and savings deposites and other deposit for transferable foreign currency, certificates as well as securities repurchase agreements) plus checks for travelers, paper for trades, time deposits for foreign currency, and share of funds for

the market It plays the important role in improving the health of the banking system

as well as whole the economy Consequently, positive sign is expected for this indicator

- The ratio of Stock-Market-Capitalization to GDP It equals the total value of all listed shares in a stock market as a percentage of GDP It is benefit to boost the economic growth, so the expected sign of the relationship between this variable and economic growth is positive (Jun, 2012; Levine & Zervos, 1996; Ndlovu, 2013; Wu

et al., 2010)

Table (1) summary information about variables measured

Variable Definition Expected sign Data sources

Rm Lagged market excess return Ambiguous Datastream International

Vol Lagged bank volatility Negative Datastream International

Indicators of country characteristics

Political Political stability and

Absence of violence

Indicators of financial development

Stock_cap Stock market capitalization

to GDP

Positive World Bank Annually

With the purpose of prolong the time-series information in this research, this study handle yearly data by overlapping method with observations in quarter The descriptive statistics and correlation matrices for Growth (GDP growth rates), vol (banking industry volatility), Rb (lagged bank excess return), and Rm (lagged market excess return) for comparison and are presented in Table 2

3.2 Methodology

We apply the generalized_method_of_moments (GMM) econometric techniques developed for dynamic panel-data models Base on the model of Cole et al., 2008, Campello et al., 2010, Cornett et al., 2010 ; Moshirian & Wu 2012; Arellano and Bond, 1991; Arellano and Bover, 1995; Ahn and Schmidt,1995; Blundell and Bond, 1998) The difference GMM and system GMM become more popular due to its simple implementation and the weak assumption on instrument variables We examine a fixed-effect dynamic model for full sample and five subsamples at the beginning:

Yit = α +  Yi(t-1) +  ’Xi(t-1) + ni + it

In above model, i and t is used to indicate country and time period; Yit is the GDP growth rate for selected samples at the time t; Yi(t-1) is the lagged values of dependent variable X is vector of explanatory variables, they contain lots of variables, such as the primary variable banking industry volatility (VOLit), the controlled variables Lagged market excess_return (Rm), and the interaction terms between banking industry volatility and the variable of the banking institutional characteristic and country specific Where, ni is the unobserved specific effect for

country i, it is an error term We use these interaction terms to examine the effects of indicators of country characteristic factors and indicators of financial development on economic growth in interacting with bank volatility In the process

of applying GMM method, we eliminated the group effect from the fixed-effect model by one simple technique, which is take the first-difference Consequently, we have:

The endogeneity in above model cause serious problems, such as results calculated are inconsistent and bias, the link between the lagged dependent variable and the error term To handle these problems, we use properly instrument variables as suggesting of Arellano and Bond (1991): “instruments are lagged values of explanatory-variables in the regression at the original level” The authors also give out the assumption that the link between disturbances in the time-varying does not exist

εit, E(εitεis) = 0, for i = 1, , N and ∀t / = s; and the initial conditions Yi1 is not

correlated with future realizations of the error term, E(Yi1it) = 0, for i = 1, , N and t = 2, , T, we can use the following m = 0.5(T− 1)(T−2) moment conditions for the autoregressive parameter:

The generalized method of moments (GMM) estimators could be given by:

In model above, Wi is the (T − 2)×q (q is the number of regressors) matrix, Zi is the (T − 2) ×m matrix, yi is the (T − 2) vector, and AN is the weighting matrix Here, the choices of AN give rise to a set of GMM estimators based on the moment conditions The difference GMM is called original estimator However, this first- differenced estimator is less suitable for reducing the sample length, surveying huge amount of information in the level of the variables, and in the indirect-link among the levels and the first differences Thus, we will have the inefficient calculation (Ahn and Schmidt, 1995) System GMM will have lower bias, high precision result

in finite sample, so it is introduced by Arellano and Bover (1995) to handle these problems In the system GMM, the original level is linked to the first-differenced regressions In the specific way, the instruments in the level regressions are the lagged first-differences variables, and instruments in the first-differenced regression are the lagged level variables Hence, we will get the original level regressions, then

we have the additional moment conditions are as follow:

In our study, there are two main techniques employed for panel data: the differenced GMM method, GMM(DIF), and the updated one as system GMM, GMM(SYS) GMM method provide consistent estimators, so they has a lot of advantages to other method when applying for the dynamic panel Although Blundell and Bond, (1998), and Arellano and Bover, (1995) prove that the GMM(SYS) method has some weak points compared to GMM(DIF) one, particularly, in the case of low autoregressive parameter, less than 0.8, and large number of time-series observations The data of this research comprise 22 economies The shortest time-series observation has 07 quarter, and the longest one has 47 quarter In general, the autoregressive parameter is small, and the number of time periods is large Therefore, the results of measuring in both GMM(SYS) and GMM(DIF) methods are given out

first-In order to test the consistency of two GMM techniques above, two diagnostic tests are used Firstly, Arellano-Bond test is applied to check second order autocorrelation in the first differenced residuals Another kind of test that is J- statistic-Hansen test, it is applied to check over-identifying-restrictions Its performance is to examine the suitability of the model The null hypothesis of the J statistic implies that instruments are endogenous This mean that, these instrument variables are correlated with the error term In the case that the null hypothesis is rejected, the instruments satisfy the requirement of orthogonality conditions

CHAPTER 4: RESULTS AND FINDINGS

This section will be divided into two fields The first field illustrates descriptive

statistics of variables and its economic meanings The next field shows out

Econometric results and its economic meanings This field also give out the best

models chosen, and the unsuitable models uncollected

4.1 Descriptive statistics of variables

Summary descriptive statistics for the panel data are presented in Table 2

Table 2

Summary descriptive statistics of primary variables

All economies Upper middle income Low and Lower middle income

Africa South Asia & East Asia Latin America

Source: World Bank

Share of three subsamples by geographical region

Table 3 : Country specifics

Economy Region Sample period Interest rate(Year) Year No.of

Argentina Latin America Q2/2003 - Q4/2014 10.20 22.26 2014 6 Botswana Sub-Saharan Africa Q2/2009 - Q4/2014 4.76 12.41 2003 3 Brazil Latin America Q2/2003 - Q4/2014 11.60 22.11 2003 21 Chile Latin America Q2/2003 - Q4/2014 6.04 6.95 2008 7 Kazakhstan Europe&Central Asia Q2/2013 - Q4/2014 7.00 7.01 2007 12 Malaysia East Asia Q2/2003 - Q4/2014 2.95 3.60 2006 10 Mauritius Sub-Saharan Africa Q4/2008 - Q4/2014 5.43 10.95 2007 2

South

Africa Sub-Saharan Africa Q2/2003 - Q4/2014

7.15 10.85 2008 8 Turkey Europe&Central Asia Q2/2003 - Q4/2014 19.02 37.68 2003 17 Venezuela Latin America Q2/2003 - Q4/2014 12.51 15.00 2008 10

Kenya Sub-Saharan Africa Q1/2004 - Q4/2014 7.54 12.58 2012 8 Uganda Sub-Saharan Africa Q1/2007 - Q4/2013 11.51 16.04 2003 2

Ghana Sub-Saharan Africa Q2/2011 - Q4/2014 16.96 27.25 2003 7 Indonesia East Asia Q2/2003 - Q3/2014 8.61 11.80 2006 30 Morocco Middle East Q2/2003 - Q4/2014 5.38 9.50 2009 6 Nigeria Sub-Saharan Africa Q1/2010 - Q4/2014 9.85 14.79 2003 15 Pakistan South Asia Q2/2003 - Q2/2008 9.61 13.12 2011 19 Philippines East Asia Q2/2005 - Q2/2014 5.08 9.77 2004 1 Sri lanka South Asia Q2/2003 - Q4/2014 10.02 18.60 2008 13 Viet nam East Asia Q2/2007 - Q4/2013 6.64 12.35 2011 7 Zambia Sub-Saharan Africa Q1/2011 - Q4/2014 12.28 29.97 2003 3

F or the full panel comprise 22 economies, the average GDP growth rate is around 0.7% with a range of -20.1% to 12.6% Whereas, the average banking industry volatility is the highest (3.2%) with a range of 0.00381% to 197.8% The average market excess return is nearly the same the average bank excess return, around - 734.9% with a range of -3706.4% to 13.8%, conversely, the range of average bank excess return is larger from -4885% to 115% The extreme negative values for both market excess returns and bank excess returns might result from the high short-term interest rates that were dominant in the most of these markets during the sample periods

For the upper middle income economies, the average GDP growth rate accounts for lower level (0.7%) with a range of -20.1% to 12% In contrast, the average banking industry volatility make up the highest degree (4.6%) with a range between 0.00479% and 197.8%, the average market excess return is nearly the same the average bank excess return around -722.8% with a range of -3706.4% to 13.8%, and with a huge differences from -4885% to 115% for the average bank excess return

On the contrary, in the low income and lower middle income economies, the average GDP rate is about 0.6% with a range of -7.5% to 12.6%, the average banking industry volatility is 1.4% It have a range between 0.00381% and 63.4% Conversely, the value of the average market excess return and the average bank excess return are almost equal It accounts for the lowest level approximately - 749.9% with a very large range from around -2583.6% to around 10.7%

For the Africa economies, the average GDP growth rate is around 0.7% with a low range of -8.7% to 10.4%, the average banking industry volatility is highest (1.7%) with a range of 0.00574% to 63.4%, the average market excess return make up lowest degree (-811.9%) with a range between -2583.6% and -72.5%, and the average bank excess return is the higher (-80.56%) with a range from -2577.6% to approximately 0%

For the South Asia and East Asia economies, the average GDP growth rate is around 0.6% with a range between -7.3% and 12.6%, the average banking industry volatility is around 1% with a range from 0.00479% to 22.6% The value of the

average market excess return and the average bank excess return are almost equal

It is approximate -645.5%, and having range from about -1877% to about 2%

For the Latin America economies, the average GDP growth rate is around 0.7% with a range of -4.2% to 7.4%, the average banking industry volatility is around 7.4% with a range of 0.00975% to 197.8% The values of the average market excess return and the average bank excess return are nearly the same It account for the lowest degree (-749.9%) with a range from around -2583.6% to around 10.7%

In summary, the average GDP growth rates for the six samples are nearly the same around 0.7% with the smallest range Whereas, the value of the average banking industry volatility is positive and higher with larger range In contrast, the values of the average market excess return and the average bank excess return are negative and fluctuated widely among samples They have the largest range in general

The simple correlation between GDP growth rates and banking volatility is negative among samples, which is -1.5%, -1.6%, -3.6, -4%, -1.9%, -2.4% for all economies, upper middle income economies, low income and lower middle income economies, Africa economies, Latin America economies, respectively Nevertheless, banking volatilities are slightly correlated with market excess returns, with the simple correlation of 0.064, 0.098, -0.151 for the sample of all markets, upper middle income markets, low income and lower middle income markets, respectively This correlation is slightly higher in geographic region groups, with the simple correlation of -0.115, -0.219, 0.198 for the sample of Africa markets, South Asia and East Asia markets, Latin America markets, respectively

The financial shock in each group or each country may cause very high inflation rate in economies surveyed after that, and high interest rate simultaneously

4.2 Econometric results:

The main objective of this study is to examine whether there is the relationship between banking industry volatility and economic growth in one main sample and

five subsamples We address this issue by looking at the significance of the coefficients of relevant variables rather than the scale of the relevant coefficients

In the first stage, we carry out the GMM-Dif and GMM-Sys estimations for the all panel comprising 22 market sample In the second stage, we repeat estimations of each of the two methods using 11 upper middle income economies, 11 low income and lower middle income economies as two subsamples In the next step, we repeat estimations of each of the two methods using 8 Sub-Saharan Africa, 6 South Asia and East Asia, 5 Latin America as three subsamples

In this section, we test the ability of banking industry volatility to predict economic growth in a panel analysis combining data for all 22 markets, and then examine whether this relationship is influenced by a series of country-specific characteristics

of each country To observe this effect, we interact banking industry volatility with these country characteristics variables The signs on the coefficients of these interaction terms is the evidence for us to recognize whether these variables strengthen or weaken the impact of banking industry volatility on the economic activities

Panels A and B of Table 4 report the results for all 22 markets using difference GMM and system GMM technique, respectively Panels A and B of Tables 5 present the difference GMM results, and Panels A and B of Table 6 present the system GMM results for upper middle income markets, and low income and lower middle income markets, respectively Panels A, B and C of Tables 7 present the difference GMM results, while Panels A, B and C of Table 8 present the system GMM results for Sub-Saharan Africa group, South Asia and East Asia group, Latin America group, respectively

t statistics in parenthese: * significant 10%, ** significant 5%, *** significant 1%