EFFECT OF MACROECONOMIC VARIABLES ON STOCK MARKET VOLATILITY IN KENYA EVANS OMBIMA AMATA DOCTOR OF PHILOSOPHY (BUSINESS ADMINISTRATION FINANCE) JOMO KENYATTA UNIVERSITY OF AGRICULTURE AND TECHNOLOGY 2017 Effect of Macroeconomic Variables on Stock Market Volatility in Kenya Evans Ombima Amata A Thesis Submitted in Partial Fulfilment for the Degree of Doctor of Philosophy in Business Administration, Finance in the Jomo Kenyatta University of Agriculture and Technology 2017 ii DECLARATION This thesis is my original work and has not been presented for a degree in any other University. Signature ……………………………. Date ………………………………. Evans Ombima Amata This thesis has been submitted for examination with our approval as University Supervisors Signature …………………………….. Date ………………………………… Prof . Willy Muturi JKUAT, Kenya Signature ……………………………. Date ………………………………… Dr. Martin Mbewa Centre for parliamentary studies, Kenya iii DEDICATION I dedicate this work to my loving parent. iv ACKNOWLEDGEMENT I give all the glory and honor to God for bringing me this far. I wish to express my sincere gratitude to Prof. Willy Muturi for his inspiring guidance, scholarly interpretations and valuable criticisms. I wish to thank Dr. Martin Mbewa for his invaluable contributions to this work. I thank my family, for their prayers and inspiration. I appreciate everybody who contributed to this work. God bless you. v TABLE OF CONTENTS DECLARATION...................................................................................................................... ii DEDICATION......................................................................................................................... iii ACKNOWLEDGEMENT ...................................................................................................... iv TABLE OF CONTENTS ........................................................................................................ v LIST OF TABLES ................................................................................................................ viii LIST OF FIGURES ................................................................................................................ ix LIST OF APPENDICES ......................................................................................................... x ACRONYMS AND ABBREVIATIONS ............................................................................... xi DEFINITION OF TERMS................................................................................................... xiv ABSTRACT ........................................................................................................................... xvi CHAPTER ONE ...................................................................................................................... 1 INTRODUCTION.................................................................................................................... 1 1.1 Background of the Study ................................................................................................. 1 1.2 Statement of the Problem ................................................................................................. 5 1.3 Objectives of the Study .................................................................................................... 6 1.3.1 General Objectives .................................................................................................... 6 1.3.2 Specific Objectives ................................................................................................... 6 1.4 Hypotheses ....................................................................................................................... 7 1.5 Significance of the Study ................................................................................................. 7 1.6 Scope of the Study ........................................................................................................... 9 1.7 Limitations of the Study................................................................................................... 9 CHAPTER TWO ................................................................................................................... 11 LITERATURE REVIEW ..................................................................................................... 11 2.1 Introduction .................................................................................................................... 11 vi 2.2 Theoretical Review ........................................................................................................ 11 2.2.1 Arbitrage Price Theory (APT) ................................................................................ 11 2.3 Conceptual Framework .................................................................................................. 26 2.4 Review of Empirical Literature ..................................................................................... 28 2.4.6 Relationship between Investor Herding Behavior and Stock Market Volatility .. 40 2.5 Critique of the Existing Literature ................................................................................. 43 2.6 Summary ........................................................................................................................ 44 2.7 Research Gaps ................................................................................................................ 45 CHAPTER THREE ............................................................................................................... 47 RESEARCH METHODOLOGY ......................................................................................... 47 3.1 Introduction .................................................................................................................... 47 3.2 Research Design............................................................................................................. 47 3.3 Research Paradigm......................................................................................................... 48 3.4 Population ...................................................................................................................... 48 3.5 Sampling Frame ............................................................................................................. 49 3.5 Sample and Sampling Technique................................................................................... 49 3.6 Data Collection Instruments .......................................................................................... 49 3.7 Data Collection Procedure ............................................................................................. 50 3.8 Preliminary Data Analysis ............................................................................................. 51 3.9 Data Processing .............................................................................................................. 53 3.10 Operationalization of Variables ................................................................................... 55 3.11 Summary ...................................................................................................................... 62 CHAPTER FOUR .................................................................................................................. 63 RESEARCH FINDINGS AND DISCUSSION .................................................................... 63 4.1 Introduction .................................................................................................................... 63 4.2 Response Rate ................................................................................................................ 63 4.4 Preliminary Tests ........................................................................................................... 64 4.4 Descriptive Statistics ...................................................................................................... 69 vii 4.4.2 Interest Rate ......................................................................................................... 70 4.4.5 Investor Herding behavior ...................................................................................... 73 4.4 Correlation Analysis ...................................................................................................... 77 4.5 Stationarity and Unit Root Test ..................................................................................... 79 4.6 Cointegration Test .......................................................................................................... 81 4.7 VECM Causality Test Results ....................................................................................... 83 4.9 Granger Causality Tests ..................................................................................................... 89 4.10 Hypotheses Testing ...................................................................................................... 91 CHAPTER FIVE ................................................................................................................. 100 SUMMARY, CONCLUSIONS AND RECOMMENDATIONS ..................................... 100 5.1 Introduction .................................................................................................................. 100 5.2 Summary of Findings ................................................................................................... 100 5.3 Conclusion ................................................................................................................... 107 5.4 Recommendations ........................................................................................................ 108 REFERENCES ..................................................................................................................... 111 APPENDICES ...................................................................................................................... 139 viii LIST OF TABLES Table 4. 1: Multicollinearity Results for the Macroeconomic Variables ................................ 65 Table 4. 2: VEC Residual Portmanteau Tests Results for Autocorrelation ........................... 66 Table 4. 3: VEC Residual Serial Correlation Lagrange Multiplier Test Results .................... 66 Table 4. 4: Normality Test Results ......................................................................................... 67 Table 4. 5: Summary of Descriptive Statistical for Secondary Data ..................................... 69 Table 4. 6: Correlation Analysis ............................................................................................. 78 Table 4. 7: Unit Root Test Results .......................................................................................... 81 Table 4. 8: Cointegration Test Results .................................................................................... 82 Table 4. 9: Vector Error Correction Model Results for Model I ........................................... 86 Table 4. 10: Vector Error Correction Model Results for Model II ......................................... 87 Table 4. 11: Granger Causality Test Results (Block Exogeneity Wald Tests) ...................... 91 Table 5.1: Summary of Research Findings and Implications ............................................... 104 ix LIST OF FIGURES Figure 4. 1: Histogram of Stock Market Volatility ................................................................. 67 Figure 4. 2: Histogram for the Interest Rate ........................................................................... 68 Figure 4. 3: Histogram for Inflation Rate ............................................................................... 68 Figure 4. 4 :Histogram for Foreign Exchange Rate ................................................................ 68 Figure 4. 5: Histogram for Gross Domestic Product .............................................................. 69 Figure 4. 6: Average Monthly Exchange Rate from January 2001 to December 2014 .......... 70 Figure 4. 7: Trend of the Average Monthly T-bill Rate from January 2001 to December 2014 .......................................................................................................................... 71 Figure 4. 8: Average Monthly Inflation Rate from January 2001 to December 2014 ............ 72 Figure 4. 9: Extrapolated Monthly Average for the Gross Domestic Product ....................... 72 Figure 4. 10: Monthly Herding Index Trend from January 2001 to December 2014 ............. 73 Figure 4. 11: Trend of NASI Index from January 2001 to December 2014 ........................... 75 Figure 4. 13: Stock Market Volatility Trend from January 2001 to Dec 2014....................... 77 Figure 4. 14: Inverse Roots of the Autoregressive Characteristic Polynomial of the Estimated VAR model ...................................................................................................... 88 x LIST OF APPENDICES Appendix 1: Letter of introduction ........................................................................................ 139 Appendix 2: NACOSTI research Permit ............................................................................... 140 Appendix 3: VECM results for model I ................................................................................. 141 Appendix 4: VECM Results for Model II.............................................................................. 144 Appendix 5: Estimated Residuals of the VAR MODELS ..................................................... 154 Appendix 6: Auto correlation test results .............................................................................. 155 Appendix 7: Companies Listed on the Nairobi Securities Exchange as at December 2014 . 157 Appendix 8: Secondary Data Collection Schedule ................................................................ 161 Appendix 9: Inverse Root of AR Characteristic polynomial ................................................. 164 Appendix 10 : Summary of Empirical Literature on Relationship between Macro-economic Variables and Stock Market Volatility................................................................ 165 Appendix 11: Summary of Empirical Literature on Effect of Investor Herding Behavior on Stock Market Volatility. ...................................................................................... 175 xi ACRONYMS AND ABBREVIATIONS ADF Augmented Dickey Fuller AIC Akaike Information Criterion AIMS Alternative Investment Market Segment APT Arbitrage Pricing Theory AR Autoregressive CAPM Capital Asset Pricing Model CBK Central Bank of Kenya CMA Capital Markets Authority CPI Consumer Price Index CSAD Cross Sectional Absolute Deviation CSSD Cross Sectional Standard Deviation ECT Error Correction Term EMH Efficient Market Hypothesis FEX Foreign Exchange Rate GARCH Generalised Autoregressive Conditional Heteroskedasticity GDP Gross Domestic Product HI Herding Index INF Inflation xii INT Interest Rate IPO Initial Public Offerings JB Jarque-Bera test KNBS Kenya National Bureau of Statistics LFEX Logarithm of Foreign Exchange Rate LGDP Logarithm of Gross Domestic Product LM Lagrange Multiplier LSV Lakonishok Shleifer and Vishny MEV Macroeconomic Variables MIMS Main Investment Market Segment MPT Modern Portfolio Theory NASI The Nairobi Securities Exchange All Share index NSE Nairobi Securities Exchange NSE20 Nairobi Securities Exchange 20 Share Index OLS Ordinary Least Squares PP Phillips- Perron PS (2006) Patterson & Sharma (2006) PVM Present Value Model SC Schwarz Criterion xiii SMV Stock Market Volatility STATA Statistics and Data General Data Analysis Software TBILL Treasury Bills VAR Vector Auto regression VECM Vector Error Correction Model VIF Variance Inflation Factor WFE World Federation of Exchanges xiv DEFINITION OF TERMS Herding : It is the average tendency of a group of investors to buy (or sell) particular stocks at the same time, relative to what would be expected if investors traded independently (Chiang et al. 2010). Intentional herding: Intentional herding is more sentiment-driven and involves imitating other market participants, resulting in simultaneous buying or selling of the same stocks regardless of prior beliefs or information sets (Kremer, 2012). Market wide herding: A form of herding arising when investors in the market ignore the individual characteristics of stocks and instead follow the performance of the market (Henker, J. Henker, G & Mitsios, 2006) Stock Market Volatility: Stock market volatility is the fluctuation in the price of broad stock market index over a defined period. It is the dispersion and not the direction of changes in price (Ambrosio, 2007). Unintentional herding: Unintentional herding occurs when investors are attracted to stocks with certain characteristics such as higher liquidity, (Falkenstein, 1996), or when investors rely on the same factors and information, leading them to arrive at similar conclusions regarding individual stocks (Hirshleifer , Subramayan &Titman, 1994). xv Volatility: Volatility is the relative rate at which the price of a security moves up and down within a very short period of time (Taylor, 2007). xvi ABSTRACT Stock market volatility is widely regarded as one of the factors that erode investor confidence in African markets. This happens when a sharp fluctuation in share prices is not explained by changes in fundamental economic factors. Theories in finance have for long viewed macroeconomic variables as predictors of stock market volatility, while studies in behavioral finance have associated stock market volatility with investor behavior, particularly the herding behavior. This study sought to examine the relationship between macro-economic variables and stock market volatility in Kenya. Specifically, the study examined the direct relationship between each of the four selected macro-economic variables namely; interest rates, inflation rate, foreign exchange rate, gross domestic product, and stock market volatility. The study further explored the moderating effect of investor herding behaviour on the direct relationship between selected macro- economic variables and stock market volatility. The study adopted a descriptive research design and targeted all companies listed on the Nairobi Securities Exchange from January 2001 to December 2014. The study used secondary data on interest rate, exchange rate, inflation rate and GDP, covering a period of 14 years. The data was obtained from the Kenya National Bureau of Statistics and the Central Bank of Kenya. Data on share prices and market indices was acquired from the Nairobi Securities Exchange. Stock market volatility was measured by computing the standard deviation of the Nairobi Securities Exchange daily and monthly returns over the 14 year study period. The study used a market-wide herd index which was calculated using the Cross Sectional Standard Deviation (CSSD) method. Data was analyzed using E-views version 8. The study employed both correlation and regression analysis. Results from correlation analysis found that there was a significant relationship between all selected macro-economic variables and stock market volatility. However, when the long run and short run causal relationship was tested using vector error correction model (VECM) and xvii granger causality test, the study found that interest rate and inflation granger cause stock market volatility both in the short run and long run in Kenya, while GDP and exchange rate did not have a direct causal relationship with stock market volatility. The study also established that investor herding behaviour had no direct causal relationship with stock market volatility, however, investor herding behaviour was found to significantly moderate the relationship between exchange rate and stock market volatility on the Nairobi Securities Exchange. The study findings were limited to selected macro-economic variables and methods used in measuring and analysing the relationship. Further studies are recommended to investigate other macro-economic variables in order to understand their effect on stock market volatility in Kenya. The study recommends a strict monetary policy and control of factors contributing to change in inflation and interest rates which the study finds to be the key variables contributing to stock market volatility. 1 CHAPTER ONE INTRODUCTION 1.1 Background of the Study Stock market volatility is a critical phenomenon facing emerging markets today. It is associated with diminished investor confidence, mispricing of shares and reduced participation in the market. Porteba (2000) posits that a volatile stock market weakens investor confidence and drives down investor spending. Daly (1999) also notes that volatility of stock markets erodes confidence in the capital market when sharp fluctuation in share prices is not explained by changes in fundamental economic factors. According to Karolyi (2001), excessive stock market volatility undermines the usefulness of stock prices as a measure of the true intrinsic value of the firm. Volatility of stock markets is mostly associated with unstable macroeconomic environment which is a common manifestation among emerging markets. Mollah and Mobarek (2009) observe that emerging markets are highly volatile due to unstable micro economic environment. In the wake of the global financial crisis of 2007, and the effect it had on the global economy, policy makers and investors have increasingly sought to understand factors that affect proper functioning of stock markets. The WFE (2008) report identifies stock market volatility as one of the factors. Forgha (2012) observes that stock market volatility has attracted immense interest among economists, stock market analysts, government regulatory and policy makers. Policy makers and practitioners are interested in understanding the causes and possible remedies for volatility of stock markets which is singled out as one of the main reason for the underperformance of African markets. As Nyong (2005) observes, the interest in volatility of stock markets is driven by its implications to economic growth. Orabi et al. (2015) posits that policy makers are interested in the main determinants of volatility and 2 its spillover effects on real activities. This study therefore, seeks to examine the effect of macro-economic variables on stock market volatility in Kenya.. 1.1.1 Stock market volatility in Kenya In Kenya, the sessional paper No. 10 of 2012 on Vision 2030 identifies market volatility as a major challenge facing the Nairobi Securities Exchange. The Kenya financial sector stability report (2010), reports that the Nairobi Securities Exchange witnessed volatility from 2008 through 2010. According to NSE (2011), the Nairobi Securities Exchange witnessed drastic volatility in the last six months of 2011. During this time, the NSE 20 share index recorded a variance from a high of 4495 points to a low of 3733 points with market capitalization declining from Sh1192.28 billion to Sh1049.56 billion. According to Corradi, Distaso and Mele (2006), understanding the origins of stock market volatility has long been a topic of considerable interest to policy makers and financial analysts. Orabi and Algurran (2015) affirm that policy makers are interested in the main determinants of volatility and its spillover effects on real activities. By understanding the determinants of stock market volatility, policy makers will be able to forecast possible trends in the market and manage the risk facing market players. Corradi et al. (2006), posits that predicting stock market volatility constitutes a formidable challenge but also a fundamental instrument to manage the risk faced by investors. 1.1.2 Stock Market Volatility and Macro-Economic Variables Stock market volatility has for long been linked to a number of factors ranging from macro-economic to behavioral. Caner and Onder (2005), outline sources of stock market volatility as dividend yield, exchange rate, interest rate, inflation rate and movement of world market index. Abugri (2008), Caner and Oder. (2005), and Granger, Hwang and Young (2000) identifies inflation rate, interest rate, exchange rate, dividend yield and money supply as notable factors influencing stock market volatility. 3 Theories in finance have strongly linked stock market volatility to changes in macro-economic variables. One such theory is the Arbitraged Pricing Theory (APT) and the Efficient Market Hypothesis (EMH). The APT, suggested by Ross (1976) and efficient market hypothesis by Fama (1970) have for long been used by researchers to explain the relationship between changes in macro-economic variables and stock market volatility. The efficient market hypothesis holds that prices adjust rapidly to new and relevant price sensitive information (Dowling, 2005). Part of the sensitive information that prices would adjust to according to the EMH is information on changes in macro-economic variables such as interest rate, inflation and exchange rate. The theory complements the philosophy underlying fundamental analysis which suggests that the intrinsic value of a security is partly determined by the underlying economic variables. The APT, developed by Stephen Ross in 1976 as an alternative to the capital asset pricing model (CAPM), explains the relationship between return and risk and relates the expected return of a share to the return from the risk-free asset and a series of other common factors that systematically affect the expected return of a share (Balla, 2006). The common factors in this theory are the various macro- economic variables which affect share prices and can be a source of volatility in stock markets. However, the APT does not specify which macroeconomic variable is most responsible for stock market volatility. This leaves researchers with an open field to explore the numerous macro-economic variables and establish which of the factors has the highest prediction powers for stock market volatility. This study is therefore contributing to this endeavor by focusing on the effect of interest rate, inflation, exchange rate and gross domestic product on stock market volatility. The study also explores the moderating effect of investor herding behavior on stock market volatility. 1.1.3 Investor Behavior and Stock Market Volatility The emergence of behavioral finance has brought about a paradigm shift from the traditional finance theory in explaining certain occurrences in stock markets. 4 Behavioral finance associates stock market volatility to the behavior of investors in the market rather than economic fundamental. According to Shiller (2000), stock market volatility is due to fundamental shift in investors‘ behavior. Shiller (2000) observes that a shift in investor behavior is driven less by economic fundamental and more by sociological and psychological factors. Proponents of behavioral finance have identified a number of behavioral biases as key in explaining stock market volatility. One of the most cited behavioral bias is the herding behavior. Shefrin (2000) notes that price adjustments are not only due to the arrival of new information but also due to market conditions or collective phenomena such as herding. This affirms the effect that herd behavior has on stock markets. Tan, et al., (2008) also posits that the influence of investor herds‘ drives prices away from their fundamental values. According to Christie and Huang (1995), investor herds are frequently used to explain stock market volatility. This study therefore explores the effect of investor herd behavior on the Kenyan market. The study draws guidance from a number of behavioral finance theories. Among the behavioral theories that underpin the relationship between investor herd behavior and stock market volatility are; information cascade theory and prospect theory. The information cascade theory postulates that individuals make decisions based on observation of others without regard to their own private information (Hirshleifer, 2001). The prospect theory proposes that people are not rational and therefore do not make investment decisions based on economic fundamentals but rather on psychological factors (Kahneman & Tversky, 1979) 1.1.4 Investor Herding Behavior in Kenya Herding takes place when investors mimic others ignoring their own substantive private information (Sias, 2004). This is a common phenomenon in emerging markets, due to the capital market environment which is said to encourage the manifestation of herd behavior (Gelos & Wei, 2002). According to Kallinterakis (2007), some of the factors that promote herding behavior in emerging market are; 5 information asymmetry, feedback trading, institutional risk management systems, market manipulation and size of firms listed on the securities market. A few studies have investigated the relationship between stock market volatility and herd behavior in Kenya and confirmed the presence of herd behavior. Kimani (2011), Werah (2006), Aduda and Muimi (2011), (2012) and Yenkey (2012) found that investors on the Nairobi Securities Exchange are influenced by behavioral biases; key among them is the investor herd behavior. Werah (2006) found that the behavior of investors at the NSE is to some extent irrational in regard to fundamental estimations as a result of anomalies such as herd behavior. Yenkey (2012) studied how nascent investors invested on the Nairobi Securities Exchange following initial public offerings and found that newly recruited investors who joined the Securities through IPOs, presented herd-like behavior by mimicking the trading behavior of experienced institutional investors. Findings in these studies affirm that herding is a common phenomenon on the Nairobi Securities Exchange. However, little is known about the effect of herding behavior on the Nairobi Securities market. This is due to the very few studies done to examine the relationship between investor herding behavior and stock market volatility in Kenya. This therefore, calls for the need to investigate the effect of investor herd behavior on the volatility of the Kenyan market. 1.2 Statement of the Problem Volatility of stock markets threatens economic growth and efficient allocation of resources. According to Daly (1999) volatility of security markets erodes confidence in the capital market, reduces liquidity and discourages wide participation. The sessional paper No. 10 of 2012 on Kenya Vision 2030 highlights market volatility as one of the leading problems facing the Nairobi Securities Exchange. According to the financial sector stability report, (2010), the Nairobi Securities Exchange witnessed volatility in 2008 through 2010, during this time; the volatility index stood at 56.93, rose to 150.16 in March 2010 and dropped to 67.84 in June 2010. According to the NSE report, (2011), the NSE 6 witnessed drastic volatility in the last six months of 2011 where the NSE 20 share index recorded a variance from a high of 4495 points to a low of 3733 points. Available literature supports the link between stock market volatility, macroeconomic variables and investor herd behavior. Christie and Huang (1995) opine that investor herds are frequently used to explain market volatility. Tan, et al., (2008), records that the influence of investor herds‘ drives prices away from their fundamental values. Studies done in Kenya find significant evidence of herd behavior among investors. Wamae (2013) finds that herding influences investment decision making among investment banks in Kenya. Yenkey, (2012), finds that newly recruited investors through IPOs, present significant levels of herding behavior. Most studies on market volatility in Kenya have focused largely on macro- economic variables. A few studies have investigated the effect of herd behavior on stock market volatility. The purpose of this study, therefore, is to bridge this knowledge gap by investigating the moderating effect of investor herd behavior on the relationship between macroeconomic variables and stock market volatility in Kenya. 1.3 Objectives of the Study 1.3.1 General Objectives The general objective of this study was to investigate the causal relationship between macro-economic variables and stock market volatility in Kenya. 1.3.2 Specific Objectives The study aimed at achieving the following specific objectives: 7 1. To establish the relationship between inflation rate and stock market volatility in Kenya 2. To examine the relationship between interest rate and stock market volatility in Kenya. 3. To establish the relationship between exchange rate and stock market volatility in Kenya 4. To determine the relationship between the gross domestic product and stock market volatility in Kenya 5. To explore the moderating effect of herd behavior on the relationship between macroeconomic variables and stock market volatility in Kenya. 1.4 Hypotheses To achieve the above objectives, the study sought to test the following null hypotheses: 1. H0 There is no significant relationship between changes in inflation rate and stock market volatility in Kenya 2. H0 There is no significant relationship between changes in interest rate and stock market volatility in Kenya. 3. H0 There is no significant relationship between changes in exchange rate and stock market volatility in Kenya. 4. H0 There is no significant relationship between changes in the Gross Domestic Product and stock market volatility in Kenya. 5. H0 Herding behavior does not significantly affect the relationship between selected macroeconomic variables and stock market volatility in Kenya. 1.5 Significance of the Study This study is of significance and interest to various stakeholders. Policy makers in Kenya would greatly benefit from findings on the causes of stock market 8 volatility. Knowledge of factors causing stock market volatility is critical to enable policy makers control the direction, magnitude and stability of the economy by adjusting macroeconomic variables if the relationship between stocks returns and economic activity has predictive power to stimulate the growth of the economy. The Kenyan development plan, encapsulated in the vision 2030, aims to achieve an annual economic growth rate of 10%, with an investment rate of 30% being financed mainly from mobilization of domestic resources. Findings in this study provides the understanding of how studied macro-economic variables affect the performance of the securities exchange and will help policy makers in formulating policies to enable the government actualize the Vision 2030 dream by strengthening the capital markets and raising the requisite capital for envisioned projects. The study is likely to contribute greatly to the growing literature in behavioral finance particularly herd behavior. Empirical literature documents that emerging markets constitute environments whose institutional structures naturally facilitate the manifestation of herd behavior yet very little is known on the effect of herd behavior on stock markets of emerging economies. Understanding the relationship between stock market returns and macroeconomic fundamentals is important to both academics and policy makers. Since the extent and direction of the relationship is still inconclusive for both emerging and developed economies, this study offers a contribution to literature by examining the relationship between stock market volatility, interest rate, inflation, gross domestic product, foreign exchange rate and investor herding behavior. Financial analysts and investors have had an interested in understanding the nature of volatility patterns on shares and those events that can explain the persistence of volatility over time. Findings of this study can inform the development of better investment policies by financial analysts which would in turn improve the performance of their investments portfolios. 9 1.6 Scope of the Study This study was focused on the effect of macro-economic variables on stock market volatility in Kenya. The study selected four macro-economic variables, namely; interest rate, Gross domestic product, foreign exchange rate and inflation guided by the empirical literature. The study covered a period of 14 years, from 2001 to 2014. The choice of the period was informed by two reasons. First, the period witnessed vibrant trading and relative stability, except for the 2008 post-election violence. Second, during this period the NSE issued ten IPOs, providing an excellent setting to investigate herding behavior since IPOs are greatly associated with herd behavior. Additionally, a 14 year period assures relatively more reliable findings for a time series study. Robert (2010) notes that the longer the time series period the better and the more data the better. Smolka (1997) opine that time series have to be comparatively longer to deliver reliable results. Many similar studies in Kenya; Kirui et al. (2014), Ouma et al. (2013), Olweny et al. (2011) were conducted during this time. This consequently, presents a level setting to corroborate the findings made in this study with other studies through review of empirical literature over the same period of study. 1.7 Limitations of the Study The study focused on the effect of macro-economic variables on stock market volatility and also explored the moderating effect of investor herd behavior in the relationship. The study made a number of findings which form part of the recommendation to policy makers and industry. However, a number of limitations were encountered in the course of the study. First, the study investigated four macro-economic variables out of the many macroeconomic variables proposed by finance theory, to be predictors of stock market volatility. It is possible that many macro-economic variables not selected in this study have higher predictive powers 10 than those investigated in this study. Findings in this study would therefore be limited to macro-economic variables investigated by the study. Secondly, a number of methods were available to the study for the measurement of stock market volatility and investor herd behavior. The study employed monthly standard deviations of the NASI index as a measure of stock market volatility and CSSD index as a measure of investor herd behavior. This left out other methods which could give different outcomes. The Vector Error Correction Model and the Granger causality test were used to investigate the causal relationship between macroeconomic variables and stock market volatility. Methods like TGARCH and OLS, used in similar studies, may have returned different findings. Findings in this study are therefore limited to the methods adopted in measuring stock market volatility, herding behavior and in data analysis. Thirdly, the social-political environment changed from time to time due to political and social events over the study period. Such events were non-economic in nature and yet had a significant effect on variables used in this study. One of the events is the 2008 post-election violence which significantly affected market trading on the securities market. The findings in this study are therefore limited by the possible effects of such social political events on study variables. 11 CHAPTER TWO LITERATURE REVIEW 2.1 Introduction This chapter reviews the theoretical and empirical literature on the effect of macro-economic variables and investor herding behavior on stock market volatility. The first part reviews theoretical literature while the second part reviews the empirical literature on the relationship between stock market volatility, herding behavior and macro-economic variables. The third part provides a critique of existing literature and presents the research gap. 2.2 Theoretical Review 2.2.1 Arbitrage Price Theory (APT) The Arbitrage Pricing Theory developed by Stephen Ross in 1976 describes how financial assets are priced given the risk associated with them (Alshogeathri, 2011). The theory proposes that share prices are driven by multiple macro- economic factors (Dincer, 2014). The APT predicts that any anticipated arrival of new information about, exchange rates, interest rates, inflation, GDP, and many other macroeconomic variables will alter share prices through the impact they have on expected return (Chinzara, 2010). This theory explains how changes in macro-economic variables would influence rapid fluctuations in share prices or stock market volatility. The theory has been used in a number of studies to explain the relationship between macro-economic variables and stock market volatility. 12 Alshogeathri (2011) and Okorafor (2008) are among studies that used the APT to explain the relationship between macroeconomic variables and stock market returns. Amos (2010) studied the APT and empirical evidence in the Nigeria capital market and found that amongst the five macroeconomic variables examined, none of the variables was significant enough to stimulate the stock returns. Arewa, et al. (2013) conducted a study to test the APT on Nigerian stock market, and made findings that provided overwhelming evidence in support of the APT pricing model as a good description of expected return. According to the theory, the expected return of a financial asset can be modeled as a linear function of various macroeconomic variables or theoretical market indices, where the sensitivity to change in each factor is represented by a factor specific beta coefficient (Gay, 2008). Elton et al. (2011), opines that the APT can be tested over a class of assets such as common stocks or a small set of stocks that form the stock market index. As a single-factor model, uncertainty in asset returns comes from a common macroeconomic factor and a firm-specific cause, where the common factor has zero expected value (McMenamin, 2005). The APT can be mathematically expressed as (Kevin, 2015); …………………………………………………. (1) Where; is the return of the stock i at time t, is the risk free interest rate or the expected return at time t, is a vector of the predetermined economic factors or the systematic risks and, 13 is a measure of the sensitivity of the stock to each economic factor included in is the error term representing unsystematic risk or the premium for risk associated with assets that cannot be diversified. The one-factor model can be extended into a multifactor model by allowing for other factors that might affect stock returns by affecting its risk (Gibson, et al., 2010). The model may be modified to incorporate interest rates, inflation, gross domestic product and foreign exchange rate as specified in this study. The Arbitrage Pricing theory fails to specify the type or number of macro- economic variables to be included in studies (Fabozzi, 2015). Consequently, researchers have examined various factors in attempt to explore factors that influences stock market returns to a great extent. Ross, et al. (1987) examined the effect of inflation, gross domestic product, investor confidence, and the shift in the yield curve on stock market returns. Fifield, et al., (2002) endorses GDP, inflation, money supply and short-term interest rates as most suitable macro- economic variables for research in emerging markets. The theory was key to this study in explaining the effect of macro-economic variables on stock market volatility. 2.2.2 The Efficient Market Hypothesis The Efficient Market Hypothesis developed by Fama (1965, 1970) explains the notion that share prices in the securities market reflect all available information such that traders in the market cannot be able to make abnormal profit regardless of the level of information they possess. The hypothesis holds that prices adjust rapidly and biasedly to new and relevant price sensitive information (Moolman & Toit, 2005). According to Fama (1970), an efficient market is one in which prices always ―fully reflect‖ available information (Pilbeam, 2010). The theory envisages a perfect market where, all the available information regarding a stock‘s 14 return and risk are factored into the market price (Westbrook, 2014). It assumes that, stock prices will only be influenced by news or information (Shiller, 1997). Consequently, when any relevant information becomes available, stock prices will move immediately to reflect the new situation (Malkiel, 2003). The EMH is anchored on the assumptions that, a large number of profit- maximizing investors operate independent of each other, new information regarding securities comes to the market in a random fashion whose announcement overtime is generally independent of one another and, investors adjust security prices rapidly to reflect the effect of the new information (Lee et al., 2009). The theory explains the effect that new information has on share prices and market return. The sensitivity of such information would have an effect on share prices and market return. Hameed and Ashraf (2006) posit that increase in volatility can be attributed to absorption of new information. Karmaka (2006) opines increased volatility which is not explained by fundamental economic factors, tends to cause share prices to be mispriced leading to misallocation of resources. The theory has therefore been able to explain the effect of changes in macroeconomic variables on stock market return and volatility. According to Marx et al., (2003), the EMH postulates three forms of market efficiency namely; weak form, semi-strong form and strong form. The weak form of market efficiency asserts that asset prices incorporate all relevant past information, such as past asset prices, security dividends, and trading volume (Madura, 2008). The semi-strong form of market efficiency maintains that all publically available information is fully reflected in security prices. Publically available information includes past asset prices, company performance, political news, publicly available analysis or projections and information about expectations of macroeconomic factors. The strong form market efficiency states that stock prices reflect all relevant information, including information only known to company insiders (Ranganatham & Madhumathi, 2006). In the strong form market efficiency, all 15 market participants can freely access all available information relevant to forming opinion about the price of a security and no group of investors has monopolistic access to such information as to make abnormal profit. The efficient market theory assumes that asset prices evolve in a random walk fashion. In this theory, asset prices cannot be predicted, suggesting that investors cannot beat the market. However, a number of studies have provided evidence showing that asset prices can be predicted. Shiller (1984) and summers (1986) found that share prices and returns are predictable. Nyong (2005) based on stock returns in Nigeria, South Africa and Brazil rejected the random walk hypothesis implying that share prices are predictable. The theoretical foundation of the EMH is based on three key propositions. First, investors are assumed to be rational and value securities rationally. Second, in case some investors are irrational, their trades are random and cancel each other out without affecting prices. Third, rational arbitrageurs eliminate the influence of irrational investors on market (Cullen, 2014). The EMH posits that any fresh news about a security should be reflected in its price promptly and completely and prices should not move as long as there is no new information about the company, since it must be exactly equal to the value of the security. This means non- reaction to non-information (Shleifer, 2000). This position has been challenged in behavioral finance where share prices are found to change based on the influence of behavioral biases even when new information is not witnessed in the market. The relevance of the efficient market hypothesis to this study is twofold. First, Fama (1970) define a market as efficient if prices always fully reflect all available information. Part of the publicly available information would be a change in the economic fundamentals including; interest rates, inflation, foreign exchange rate and GDP. This theoretical foundation therefore, provides an underpinning for the relationship between macro-economic variables and stock market volatility in this study. 16 2.2.3 The Present Value Model According to Attari et al. (2013), the present value models (PVM) using future expected earnings and future expected discount rates has been empirically tested for predicting stock prices. The model explains the dynamic relationship between stock market volatility and economic activities (Semmler, 2006). Sarkar, (2012) opines that the PVM explains the relationship between stock prices and macroeconomic variables Attari et al. (2013) posit that the PVM is useful in establishing a long term relationship among stock prices and macroeconomic variables. In the words of Shiller, (1992), the model states that the price of a share is the present discounted value of the expected future dividends. The description of this model is that the expected future dividend of company shares reflects the levels of macroeconomic activities. Volatility of share prices and stock market returns would therefore, be influenced by expected future cash flows which are a function of microeconomic variables. The model has been tested and used in a number of studies. Alshogeathri (2011), Osisanwo (2012), Sarkar (2012), Attari et al. (2013) and Oseni (2011) are among studies which have used this model to explain the effect of macroeconomic variables on stock market volatility. According to the model, the difference between the intrinsic value of a share and its market value represents an overvalued or undervalued stock. The profit opportunities represented by the existence of undervalued and overvalued stocks motivate investors to trade, and their trading moves share prices toward the intrinsic value (Gorton & Allen, 1993). Consequently, investors search for mispriced stocks and their subsequent trading make the market efficient causing shares to reflect their intrinsic values. According to Banerjee (2015) the intrinsic value of a share is the present value of the cash flows the shareholder is expected to receive. The PVM can be expressed as follows (Semmler, 2006) and (McMillan, 2010): 17 Where; is the stock price. is the expected stream of returns. are the factors associated with the discount rate of future cash flows. Factors associated with the discount rate ) are factors which directly or indirectly affect expected returns and later affect stock prices. Consequently, new related macroeconomic information may be analyzed as long as they impact the expectation of stock prices or returns, the discount rates, , or both (Alshogeathri,2011). The advantage of the present value model is that it can be used to focus on the long run relationship between the stock market returns and macroeconomic variables (Osisanwo, 2012). The present value model was important to this study in explaining the relationship between macro-economic variables and stock returns volatility. The theory relates a change in share prices to the discount rate in the market which is influenced by a change in macro-economic factors. Ackert and Smith (1993) argue that volatility in stock prices is due to either a change in the discount rate or new information concerning future cash flows received by shareholders. 2.2.4 Fisher’s Theory The Fisher‗s theory of capital and investment was introduced by Irving Fisher in his Nature of Capital and Income (1906) and Rate of Interest (1907). The theory has its clearest and most famous exposition in his Theory of interest (1930). The theory defines the relationship between inflation on one hand and real and 18 nominal interest on the other hand. The fisher theory states that nominal interest rates in two or more countries should be equal to the required real rate of return to investors plus compensation for the expected amount of inflation in each country (Dimand, 2003). Fama and Schwert (1977) explains the fisher effect theory by stating that, if the market is efficient and reflects all the available information at time t-1, the price of common stocks will get adjusted so that the expected nominal return from t-1 to t is the sum of the appropriate equilibrium expected real rate and the market‘s assessment of expected inflation rate for the same time period (Waweru, 2014). According to the fisher effect theory, shares serve as hedges against inflation because they represent claims to real assets, which suggest that a positive share price is correlated to expected inflation (Dimand, 2003). 2.2.5 Information Cascade Theory Investor herding behavior can well be explained by the information cascade theory (Bruun, 2006). Lakonishok et al. (1992), Choe et al. (1999) and Wermers (1999) contend that volatility is closely related to information-induced herding. According to Anderson (2001), the theory advances a situation where people with private, incomplete information make public decisions in sequence. The first few decision makers reveal their information to subsequent decision makers who follow an established pattern even when their private information suggests that they should deviate from that pattern. Kim et al. (2013) opines that investors who trade after a cascade has started provide no information to subsequent traders because they are merely copying the action of others. Such traders do not base their decision on any information. In general, the information cascade model is guided by the idea that individuals make decisions based on observation of others without regard to their own private information. Information cascades start in a stock market when investors ignore their own information and instead infer information from a herd (Sias, 2004). 19 Todd et al. (2009) illustrates information cascade by thinking of a person who chooses between two unfamiliar, apparently similar restaurants situated on opposite sides of a street. If it‘s assumed that a customer has heard other customer‘s mixed opinions about one of the restaurants (A) and only good things about the other restaurant (B). When approaching the restaurants, the customer notes that restaurant A is more crowded than restaurant B. According to Todd et al. (2009), many people would probably then choose restaurant A, without any proof that restaurant A is better than restaurant B. The fact that many people are in restaurant A may thus be enough to attract additional customers, even if they have opposing private information. The explanation for this behavior is that people base their decisions on choices made by others. The Information cascade theory is advanced from the Avery and Zemsky (1998) and, Bikchandani et al. (1992) models. In the two models, if an investment cascade starts in a market, a long sequence of buy or sells trades is expected. According to Bikchandani, et al. (1992) model, public information or the arrival of a highly informed investor will quickly stop an incorrect information cascade. In the Avery and Zemsky (1998) model, price adjustments make it unlikely for cascades to occur and decrease the prolonged existence of cascades. Avery and Zemsky (1998), outlines three conditions which need to be present for herding to occur namely; information asymmetry, value uncertainty and, event uncertainty or uncertainty about whether the value of an asset has changed from its initial expected value. According to Blasco (2006), the link between investor herd behavior and market volatility was first noted by Friedman (1953) who found that irrational investors destabilized prices by buying when prices were high and selling when they were low. The information cascade theory, which is key to this study, underpins the fact that investors ignore their own current private information to mimic other investors. The herd phenomenon happens when investors in the market move as a group to make similar investment decisions, pushing prices away from their 20 economic fundamentals. This act by investors results into price momentum and stock market volatility. Bikhchandani et al. (1992) posits that herding behavior leads a group of investors to move in the same direction, pushing stock prices further away from the economic fundamentals, causing price momentum and excess volatility. According to Shiller (2005), herding is a collective irrationality of investors that leads to the mispricing of economic fundamentals. 21 2.2.6 Prospect Theory The prospect theory is a behavioral finance theory for decision making under uncertainty developed by Daniel Kahneman and Amos Tversky in 1979. The theory attempts to acknowledge that investors are not rational as presumed by normative theories like the efficient market hypothesis and the expected utility theory. The theory relies on observation of what people should actually do or how they actually behave. It is based on empirical evidence that people do not behave in accordance with the normative models when it comes either to decision making or choices (Lowies, 2012). The prospect theory views the investment decision making process as driven by irrational factors like herding behavior rather than economic fundamentals as stated by normative theories like the arbitraged pricing theory and the efficient market hypothesis. To this study, the theory introduces another angle of understanding volatility of stock market where irrational behavior like herding guides the investment decision processes causing price momentum and excess volatility. The theory underpins this study by confirming that stock market volatility is not only influenced by changes in the economic fundamental but by investment decisions made based on irrational factors like herding behavior. Kahneman and Tversky (1979) demonstrated in numerous experiments that the day-to-day reality of decision makers varies from the assumptions held by economists (Goldberg & VonNitzsch, 2001). According to Bing and Jason (2004), the prospect theory can help in understanding the choices and trading behavior of investors in financial markets and to explain asset pricing ―anomalies‖ including the equity premium puzzle, momentum strategy, excess volatility, IPO under-pricing and long-term performance of IPO‘s. Babaries et al (2001) confirms that models based on the prospect theory can explain the high mean excess volatility and predictability of stock returns. 22 2.4.7 Herding Theory Herding occurs when individuals mimic others, ignoring their own substantive private information (Scharfstein & Stein 1990). It is the most common behavioral factor in decision making, where investors follow investment decisions taken by the majority. Herding is a major concern for traders and policy makers as it leads to unnecessary volatility and more frequent extreme observations (Demirer, et al., 2009). Herding can either be intentional or unintentional (Bikhchandani & Sharma (2001). Intentional herding involves imitating other market participants, resulting in simultaneous buying or selling of the same stocks regardless of prior beliefs or information sets. Intentional herding can lead to asset prices failing to react to fundamental information, exacerbation of volatility, and destabilization of markets, thus having the potential to create, or at least contribute, to bubbles and crashes on financial markets ( Morris & Shin 1999) and (Persuaded, 2000). Unintentional or spurious herding is mainly fundamental driven and arises because institutions may examine the same factors and receive correlated private information, leading them to arrive at similar conclusions regarding individual stocks (Hirshleifer et al., 1994). Kallinterakis (2007) highlights some of the factors that promote herding behavior in emerging market as information asymmetry, feedback trading, institutional risk management systems, market manipulation and size of firms listed on the securities market. Information asymmetry is about non availability of information relevant to investment decision making. Information asymmetry a common phenomenon in emerging markets due to incomplete regulatory frameworks, especially in the area of market transparency. Such environments cause deficiency in corporate disclosures and poor quality of information leading to information asymmetry. According to Gelos and Wei (2002), deficiencies in corporate disclosure and information quality create uncertainty in the market, throw doubt on the reliability 23 of public information, and as a result impede fundamental analysis. Kallinterakis (2007) argues that in such an environment it is reasonable to assume that investors will prefer to base their trading on observation of others. Accordingly, intentional herding is more likely to occur in less developed markets than in developed markets. Feedback trading occurs when investors react to information in a similar manner, without necessarily ignoring their own private information. This happens when investors naturally make investment decisions based on the information feed into the market. They sometimes react to common signals in the same way. Such common reaction leads to unintentional herding. A manifestation of this kind of herding is momentum investment. This is also be called positive feedback trading. DeLong, et al. (1990); Sentana and Wadhwani (1992) suggest that positive feedback traders buy stocks in a rising market and sell stocks in a falling market, while negative feedback traders follow an investment strategy of ―buy low and sell high.‖ According to De Long (1990), positive feedback trading may lead to unintentional herding and could have a destabilizing impact on financial markets. Risk management systems used by institutional investors are another source of herding behavior. Institutional investors make use of market sensitive risk management systems in their investment management practice. Persaud (2000) and Jorion (2002), argue that market-sensitive risk management systems used by banks, such as Value at Risk (VaR) models, require banks to sell when volatility rises. Thus, banks act like a herd, all selling the same stocks at the same time in response to negative shocks. Although this kind of trading is considered to be unintentional herding, it leads to further slumps in prices. Group conformity is a psychological manifest in human desire to conform to a majority decision. It is a major reason for herding behavior. According to Bikhchandani et al. (1992), Avery and Zemsky (1998) and Park and Sabourian (2011), rational traders copy the investment activity of other market participants because they assume that others have important information. Smith and Bern 24 (1990), find that 74% of subjects in their study change an individual opinion that appears to be correct in order to conform to group consensus. The fear of losing one‘s reputation as an advisor or institution investor, by making a wrong decision, causes most practitioners to mimic majority decisions in order to avoid reputational risk. One of the explanations for herding behavior is derived from the reputation based model originally developed by Scharfstein and Stein (1990). According to this model, institutions or professional investors are subject to reputational risk when they act differently from the crowd. An explanation for reputational herding is that failing conventionally is better for one‘s reputation than succeeding unconventionally. This is because investors who herd are able to share the blame and hide in the herd when making unfavorable investment decisions (Devenow & Welch, 1996). According to Scharfstein and Stein (1990), an unprofitable investment harms a decision maker considerably less when others have made similar investments, which constitutes a reputational reason for investors to ignore private information in favor of trading with the herd. Scharfstein and Stein (1990) observe that investors do not take contrarian positions for the fear of damaging their reputation in the labor market as sensible decision makers. As a result investors ignore their private information and follow market consensus. The state of the overall market in terms of adverse stages of a business cycles or financial crisis is a source of herding behavior. Chiang and Zheng (2010) find that herding behavior is more apparent during the period in which the financial crisis occurs. Hwang and Salmon (2004) find higher herding measures during relatively quiet periods than during periods when the market is under stress. Platev and Kanaryan (2003) studied four Central Europe markets and find strong evidence of herd influence over market volatility caused by the Asian and Russian crises. Karunanayake et al. (2010) show that both the Asian crisis and the more recent global financial crisis significantly increased the stock return volatilities across all of the four markets in their study. 25 Cross market herding happens when herding in one market is affected by herding in other markets and global stock market volatility. The correlation in herding is due to geographic proximity that produces close trading relation in the region, or to a similar cultural background with less transparency and less public information available, which would induce investors to form a correlated trading decision. Experience from the Asian crisis period indicates that herding behavior tends to display co-movements (Marais et al., 2006). Experience in recent financial crises indicates that it does not matter through which channel the volatility is transmitted, (Corsetti et al., 2005). Whenever negative news develops in a given market, it will soon be learned by participants in other markets. Beirne et al. (2009) find evidence of significant stock-return volatility spill overs from the US market to many Pacific-Basin countries. Market manipulation may also promote herding, as the actions of a group of informed traders may create the impression of a profitable opportunity, thus luring others into it (Van Bommel, 2003). Hirshleifer et al. (1994), suggests that investors receive uncorrelated private information. Few early receivers trade aggressively in the initial period, subsequently reversing their positions as the later informed traders adopt the ‗follow the leader‘ strategy (Van Bommel, 2003). Small firms are usually less transparent than big firms and have less information available to public. Lack of information causes investors to imitate other traders who are perceived to have some information on small firm. The model of intentional herding has shown that there is an inverse relationship between herding and firm size (Kallinterakis, 2007). Unintentional herding, on the other hand, is more likely to occur in larger stocks because institutions have more information in common about these stocks. According to Christoffersen and Tang (2009), herding decreases with data frequency, and that herding should be less significant in stocks with larger size and higher turnover. 26 2.3 Conceptual Framework A conceptual framework is a diagrammatic presentation of variables, showing the relationship between the dependent, independent and moderating variables. The purpose of a conceptual framework is to help the reader quickly see the proposed relationship between variables in the study (Mugenda &Mugenda, 2003). This study was based on the framework provided by the efficient market hypothesis as advanced by Fama (1970), the arbitraged pricing theory suggested by Ross (1976) and information cascade theory proposed in the Avery and Zemsky (1998) model and Bikchandani, et al .(1992) model. The frame work for this study is made up of selected macro-economic variables namely; interest rates, inflation, gross domestic product and foreign exchange rate as independent variables and stock market volatility as the dependent variable. Herding behavior is included in the framework as a moderating variable. The conceptual framework below presents the relationship. 27 Independent variables Moderating variable Dependent Variable Figure 2. 1: Conceptual Framework Stock Market Volatility  Monthly Standard deviation of the Nairobi All Share Index. Investor herd behaviour  Market- wide monthly CSSD herding Index. Foreign Exchange Rate  Effective exchange rate during the study.  The Rate at which Kenyan shillings are exchanged for one US dollar. Gross Domestic Product  Income of the country Interest rate  91day Treasury bill rate. Inflation index  Consumer Price Index 28 2.4 Review of Empirical Literature Several studies have examined the dynamic relationships between stock market volatility and macro-economic variables. Limited literature is available on the effect of investor herding behavior on stock market volatility. Majority of the studies have concentrated on developed markets. Empirical literature shows diverse findings motivating more studies to understand the relationship between stock market volatility, macro-economic variables and investor herding behavior. Studies reviewed are in two categories. The first category is made up of studies which investigated the relationship between individual macro-economic variables and stock market volatility, and the second category are studies that investigated the collective effect of a number of macro-economic variables on stock market volatility. The last part of this section reviews empirical literature in relation to the relationship between investor herd behavior and stock market volatility. 2.4.1 Relationship between Interest Rates and Stock Market Volatility Finance theory proposes that interest rates and stock price have a negative correlation (Hamrita & Abdelkader, 2011). According to Jawad and Ulhaq (2012), interest rate has a more direct effect on financial market whereby an increase in interest rate causes investors to make a change in the structure of their investment, generally from capital market to fixed income securities which leads to a drop in stock prices Zhou (1996) studied the relationship between term structure of interest rates and stock market volatility on the US Market. The study used the OLS regression method to study the relationship between interest rate and variation of share prices and stock return. The study referred to theoretical literature explaining the effect of discount rates on stock market volatility. The theory proposes that the variance of dividend-price ratios may be accounted for by changing forecasts of discount rate with some unusual characteristics. The study used data on share and bond returns together with McCulloch and Kwon (1993) data set on zero-coupon yields 29 implied by the yield curve for U.S. Treasury securities. The study found that long term interest rate had an impact on stock returns, especially in the long run. The study observed that high volatility in the stock market returns was related to high volatility of long-term bond yields. The study established that long term interest rate explains a major part of variation in stock market returns. Arango et al. (2002) examined the relationship between market returns and interest rate on the Bogota stock market in Colombia. The study used daily interbank loan interest rate and share price data from January 1994 to February 2000. The study referred to the present value model and the Gordon growth model to underpin the study. The study used the smooth transition regression (STR) approach and GARCH model to examine the behavior of share prices. The study found evidence of a nonlinear and inverse relationship between share prices on the Bogota stock market and interest rate as measured by the inter bank loan interest rate. The interbank loan interest rate is normally affected by a country‘s monetary policy. Findings in this study inferred that an increase in interest rate led to a drop in share prices although the relationship was said to be nonlinear. Ahmed (2008) investigated the relationship between aggregate economic variables and stock markets in India and found a positive relationship between interest rate and stock prices on the Indian stock market. Zafar et al. (2008) examined the effect of interest rates on stock returns and volatility on the Karachi stock exchange in Pakistan. The study used GARCH models to examine the relationship and covered the period between January 2002 and June 2006. The study found a negative significant relationship between interest rates and stock market returns. Gan et al. (2006) using Forecast Error Variance Decomposition (FEVD), find a positive relationship between interest rates and returns on the New Zealand stock market. Aroni (2011) used a multiple regression method to study factors influencing stock prices on the Nairobi securities Exchange and found that interest rates significantly affect stock prices. 30 Alam, et al. (2002), used time series and panel regression to study the relationship between interest rate and stock prices in fifteen developed and developing countries namely; Australia, Bangladesh, Canada, Chile, Colombia, Germany, Italy, Jamaica, Japan, Malaysia, Mexico, Philippine, S. Africa, Spain, and Venezuela. The study found that interest rate had a significant negative relationship with share price for all the countries. Olweny et al. (2011) studied the effect of macro-economic variables on stock market volatility on the Nairobi Securities Exchange using the TGARCH method and found that interest rate affects stock market volatility. A review of empirical literature on the effect of interest rate on share prices and stock market volatility indicates generally that majority studies, Zhou(1996), Arango et al. (2002), Ahmed (2008), Zafar et al. (2008), Gan et al. (2006), Aroni (2011), Alam et al. (2002) and Olweny et al. (2011) find that interest rate affects stock prices and market volatility. However, the direction and magnitude of the effect is varied. Whereas, Zhou (1996), Arango et al. (2002) Zafar et al. (2008) and Alam et al (2002), find that interest rate has a negative effect on share prices and market volatility, Ahmed (2008) and Gan et al (2006) find that interest rate has a positive effect on share prices and stock market volatility. It‘s therefore, safe to conclude, from the reviewed literature, that interest rate affects stock prices and volatility of securities markets. 2.4.2 Relationship between Inflation and Stock Market Volatility The general understanding of the effect of inflation on share prices and market return is that inflation brings about a general increase in prices of firm inputs causing a general increase in the cost of doing business. Businesses perform poorly as a result of increased inflation hence causing share prices to drop. The drop in prices would cause investors to shift their portfolio towards other assets. This implies that inflation has the potential of influencing a change in share prices and volatility of stock markets. 31 Finance theory confirms that inflation affects the value of shares in the stock market. Fama (1981) observes a negative relationship between inflation and stock prices. Share prices are negatively impacted by inflation due to the negative correlation between inflation and expected real economic growth. Investors shift their portfolios towards real assets if the inflation rate becomes remarkably high (Hatemi, 2009). Empirical studies on the relationship between inflation and stock market volatility are yet to arrive at a consensus. Murungi (2012) examined the impact of inflation on stock market returns and volatility using OLS estimation and GARCH techniques on the Nairobi securities Exchange. The study covered the period between July 2000 and August 2012. Findings from the study revealed a negative relationship between stock returns and inflation in Kenya. A change in inflation rate had a significant negative effect on stock market volatility. Ochieng et al. (2012) studied the relationship between macro-economic variables and stock market performance. Using regression analysis, the study found a weak positive relationship between inflation and stock market return. Ouma et al. (2014), studied the impact of macroeconomic variables on stock market returns in Kenya, using ordinary least squares and found that there was a positive relationship between inflation and stock prices. Aroni (2011), using regression analysis finds that inflation significantly affects stock prices in Kenya. Kirui et al. (2014), finds an insignificant relationship between inflation and stock prices. Issahaku et al. (2013) studied the relationship between macro-economic variables and stock market returns and found a significant long-term relationship between the two variables. Ratanapakom et al. (2007) studied the relationship between macroeconomic variables and stock prices, and find that stock prices were negatively related to inflation in the short run. 32 Majority of empirical literature reviewed with regard to the relationship between inflation rate and stock market volatility show that inflation rate affects share prices and may cause volatility of stock markets. Murungi (2012), and Ratanapakom (2007) finds a negative relationship between inflation and share prices while, Ochieng and Oriwo (2012) and Ouma et al. (2014) find that inflation rate has a positive effect on share prices. The incongruous result begs for more enquiries to endorse the relationship between inflation and stock market returns in Kenya. 2.4.3 Relationship between Exchange Rates and Stock Market Volatility The effect of changes in exchange rate to share prices and volatility of stock market is well known in finance literature. Barnor, (2014) posit that the appreciation of a local currency has a tendency to hurt exporters and, consequently shares of exporting firms become less attractive. According to Joseph (2002), exchange rate changes affect the competitiveness of firms through their impact on input and output prices as a result of their unattractiveness. The market value of a share of an export-oriented firm is likely to fall. Joseph (2002) observes that when the exchange rate appreciates, exporters will be negatively affected. This happens when an appreciation in the value of a currency causes goods and services of exporting companies to be expensive on the international market. As a result, exports will decline and result into a loss of competitiveness internationally. Consequently, company profits will decline and weaken their attractiveness in the stock market (Mlambo et al., 2013). However, empirical studies have returned diverse findings some of which contradict the theory. Asaolu and Ogunmuyiwa (2011), affirms that theoretical economists and empirical researchers have not reached a consensus on the nexus between stock market volatility and foreign exchange rate. 33 Mlambo, et al. (2013) assessed the effect of exchange rate volatility on the Johannesburg Stock Exchange in South Africa. The study used the Generalized Autoregressive Conditional Heteroskedascity (GARCH) model to establish the relationship. Monthly data from 2000 to 2010 was used in the study. The study found a very weak relationship between exchange rate volatility and stock market volatility. Aslam (2014) examined the relationship between stock market volatility and exchange rate on the Karachi stock exchange in Pakistan. The study covered the period between January 2006 and December 2012 .Using different statistical tools, the study analyzed the causal relationship between both time series and found a weak negative correlation between stock market return and exchange rate. Causality test revealed that there was a bi-directional causal relationship between stock market return and exchange rate. Ambunya (2012) studied the relationship between exchange rate movement and stock market return volatility on the Nairobi securities exchange in Kenya, covering the period between 2001 and 2011. Using regression analysis, the study found that exchange rate movements greatly affected stock market return volatility and concluded that there is a strong relationship between exchange rate movement and stock market volatility in Kenya. Muhammad and Rashid (2011) studied the relationship between share prices and foreign exchange rate in Pakistan, India, Bangladesh and Sri-Lanka. The study covered the period between January 1994 and December 2000.Using the Vector Error Correction Model and Granger causality test, the study found no short run association between foreign exchange and share prices in all the four countries. With regard to long run relationship, the study found no long-run relationship between the two variables in Pakistan and India. However, findings revealed a bidirectional long-run causal relationship between exchange rate and share prices in Bangladesh and Sir-Lanka. 34 Kadir, et al. (2011) examined the predictability power of exchange rates and interest rates on stock market volatility and return in Malaysia. The study used monthly Kuala Lumpur composite Index (KLCI) returns, 3 months Malaysia Treasury bond and monthly exchange rate of Ringgit per US Dollar from 1997 January to 2009 November. Using two models based on GARCH, the relationship between exchange rate and stock market returns were found to be negative but significant for exchange rate and insignificant for interest rate. All empirical studies reviewed in this study established that there was a relationship between exchange rate and stock market volatility. The strength and direction of the relationship was nonetheless varied. Mlambo, Mandera and Sidada (2013) and Aslam (2014) found a weak negative relationship while, Ambunya (2012) and Kadir et al (2011) returned a strong significant relationship between exchange rate and share prices. It can be concluded therefore that empirical literature is in support of a significant relationship between exchange rate and share returns albeit weak in many cases. 2.4.4 Relationship between Gross Domestic Product and Stock Market Volatility It is generally understood that the economic performance of a country affect business activities in that country. This would naturally imply that the economic performance of a country would impact on performance of firms and their share prices. Therefore GDP which is a measure of the income on a country should have an effect on share prices and stock market volatility. According to Fama (1986), Mukherjee and Naka (1995) and Ibrahim and Aziz (2003), an increase in the real GDP will affect share prices through the impact it has on corporate profit. This happens when there is an increase in the real GDP, where the expected future cash flows in a company improve causing share prices to increase. This study reviewed a few studies to understand their findings in relation to the effect of gross domestic product on share prices and stock market volatility. 35 Oseni, et al. (2011) examined the relationship between macro-economic variables and stock market volatility in Nigeria. The macro-economic variables investigated in the study were; GDP, interest rate and inflation. This study employed EGARCH technique to examine the volatility in stock market and macroeconomic variables, and used LA-VAR Granger Causality test to analyze the relationship between stocks market volatility and macroeconomic variables volatility in Nigeria for the period from 1986 to 2010 using time-series data. The study found a bi-causal relationship between stock market volatility and real GDP volatility. Findings showed absence of causal relationship between stock market volatility and volatility in interest rate and inflation rate. Attari, et al. (2013) studied the relationship between macro-economic volatility and the stock market volatility in Pakistan. The macro-economic variables investigated in the study were; interest rate, inflation, and Gross domestic product. The study applied the Exponential Generalized Autoregressive Conditional Heteroskedasticity (EGARCH). Monthly time series data of the variables for the time period from December 1991 to August 2012 was used in the study. The ADF and ARCH tests was used to check the stationarity and homoscedasticity in the data respectively. The results suggested that there was no causal relationship between GDP and stock market returns. Inflation rate had a casual effect on stock market returns. Findings revealed the existence of unidirectional relationship between stock market returns and interest rate. Kibria, Kamuaran, Arshad, Perveen and Sajid (2014) investigated the impact of macroeconomic variables on stock market returns in Pakistan. The study investigated the influence of five macroeconomic variables namely; inflation, GDP Per Capita, GDP savings, money supply and exchange rate at KSE 100 index of Pakistan. The study used the annual data of 23 years from 1991 to 2013. Descriptive analysis, correlation analysis, granger causality test and regression analysis were used to study the relationship. The regression analysis results showed that inflation, exchange rate, money supply, GDP per capita and GDP 36 savings had a positive and significant impact on stock market return (KSE 100 index) in Pakistan. Hossain and Hossain (2015) examined the relationship between share price and economic growth in the USA, UK and Japan based on quarterly data for a period of 22years from 1991 to 2012.The study used the Engle-Granger co-integration and Granger causality test to determine the relationship. Findings in the study revealed absence of relationship both in the short run and long-run between share prices and economic growth in USA and Japan. Result showed a short run relationship between share price and economic growth in the UK. In terms of causality the study found that change in share price can predict short term economic growth in the USA and UK. A change in the share price was found not to be a predictor for economic growth in Japan. Adam (2015) examined the relationship between stock prices and economic growth in Indonesia using quarterly stock price index and percentage in GDP data from 2004 to 2013. Using the general univariate causal model (LVAR), the study found that there was a significant positive relationship between share prices and GDP. This meant that an increase in the share price led to an increase in economic growth in Indonesia. Accordingly, an increase in the stock price by 1 per cent led to an increase in the growth of the Indonesian economy by 0.09 per cent. A review of empirical literature on the relationship between gross domestic product and share prices and by extension stock market volatility reveals mixed findings. Whereas Oseni et al.(2011),Kibria (2014), and Adam (2015) find a significant positive relationship between gross domestic product and share prices which extends to volatility of the stock market, Hossain and Hossain (2015) and Attari et al (2013) find no significant relationship. 2.4.5 A review of other studies with various macro-economic variables After reviewing empirical literature that examined the relationship between individual or a few macroeconomic variables and stock market volatility, this 37 section reviews studies which investigated the relationship between various or a number of macro-economic variables and stock market returns. Among the studies are; Hussin, Muhammad, Abu, Awang and et al. (2012), Hasan Tarij (2009), Emeka, Aham and Uko . (2013), Aroni (2011), Kirui, Wawire and Onono. (2014), Olweny and Kimani. (2011), Ouma and Muriu (2014) and Olweny (2010). Hussin et al. (2012) investigated the relationship between the development of the stock market and macro-economic variables in Malaysia. The variables used in the study were; industrial production Index (IPI), consumer price index (CPI), aggregate money supply (M3), Islamic interbank rate (IIR) and exchange rate of Malaysia. The study used Vector Auto regression (VAR) model to examine the relationship between the stock market and macroeconomic variables in Malaysia. In order to specify the VAR model, the study determined the variables stationarity properties or integration order. The Augmented Dickey-Fuller (ADF) and Phillips-Perron (PP) unit root tests were used to determine the variables‘ stationarity properties or integration order. To determine whether the variables, particularly the stock market index and macro-economic variables were related in the long run, the contegration test used was by Johasen (1988) and Johansen and Juselius (1990). The study also did a Granger causality test in the form of vector error correction model (VECM). Granger Causality test was performed to identify the existence and the nature of Causality relationship between the variables and the stock market return. The study found that macro-economic variables were Granger cause for the stock market return. The study used monthly data for the period from April 1999 to October 2007. Findings of the study show that the stock market returns are co-integrated with the selected macro-economic variables. The stock market returns were found to be positively and significantly related to the industrial production index and the inflation index (CPI) but related negatively and significantly with money supply (M3) and exchange rate. The relationship between the interest rate and the market return was found to be negative but insignificant. 38 Patra and Poshakwale (2006) studied the relationship between the economic variables and stock market returns on the Athens Stock Exchange in Greece. The study covered the period from 1990 to 1999. The findings revealed a short term and long term relationship between inflation, money supply and trading volume and the stock prices in the Athens stock exchange. The study found no short term and long-term relationship between the exchange rates and stock prices. The data used in this study consisted of monthly closing prices of the ASE general index, Consumer Price index (CPI), Money supply (M3), and Exchange Rate and trading volumes. The variables were tested for stationarity by employing the Augmented Dickey Fuller statistic (ADF). The ADF statistics were found to be significant for all five variables. Granger causality test was used to test the short-run relationship between stock returns and economic variables. Cointegration was tested using the Eagle and Granger (1986) and Johansen and Juselius (1990). Emeka et al. ( 2013) examined the impact of macro-economic factors on the Nigeria‘s stock market returns, using Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model and annual data for the period between 1985 and 2009. The study investigated the ability of six macro-economic variables, namely inflation, government expenditure, index of manufacturing output and, interest rates, money supply, and foreign exchange rate to predict market return. Findings of this study revealed that inflation and government expenditure had a positive and significant impact on the market return. Manufacturing output and interest rate were significant and negatively related to the market return, while money supply and foreign exchange rate had no significant influence on the stock returns. The study observed that the volatility of the Nigerian stock market was more influenced by the past volatility than economic news from previous period. The study also observed that, the time varying volatility of the Nigerian stock market was persistent i.e. it took long for the market to go through a market volatility shock. Findings in this study were contrary to other studies in as far as the inflation index is related to market returns. 39 Barrows and Naka (1994), Mukherjee and Naka (1995) find that inflation negatively influences stock market return. The study used the Augmented Dickey Fuller test to check the stationarity of the variables before carrying out regression. The study also carried out a cointegration test to examine the long run relationship between the stock market return and macro-economic variables. The study preferred GARCH model to OLS model due to inadequacy of the OLS model to analyzing data that changes through time. Hassan and El Gezery (2010) examined the effect of macro-economic variables on Egyptian stock market return across different types of industries and different levels of economic states. The macro-economic variables investigated in this study were; inflation rate, interest rate, money supply and exchange rate. The study covered the period from 1993 to 2009. Findings in the study showed that the stock market index responded positively to inflation but the coefficient was insignificant. Interest rates were found to be negatively related to the market return at 10% significance level. The exchange rate was found to be positively related to the market return at 5% significance level. Money supply was found to be positively significant to the market return. The study used M2 as a proxy for the money supply in Egypt. The Vector Autoregressive (VAR) model was used to investigate the relationship between macro-economic variables and stock market return. Before inclusion of variables in the VAR model, all the variables had to be checked for stationary, since the model requires that they be stationery. The Augmented Dickey Fuller test was employed to test for unit roots and it was found that all the variables namely share price index, money supply, exchange rate; inflation rate and interest rate were stationary on first differencing at 1 percent level of significance on the basis of the ADF. The Durbin Watson Statistics and the Granger causality test were used to test whether there is one-way or bi-directional causality between Egypt stock market return and macro- economic variables. 40 Oseni and Mwosa (2011) studied the relationship between stock market volatility and macro-economic variables volatility in Nigeria using EGARCH model. The study used LA-VAR Granger Causality test to analyze the nexus between stock market volatility and macroeconomic variables volatility in Nigeria for the periods 1986 to 2010 using time-series data. The results of the findings revealed that there exists a bi-causal relationship between stock market volatility and real GDP volatility; and there is no causal relationship between stock market volatility and the volatility in interest rate and inflation rate. Olweny et al. (2011) investigated the effect of macro-economic factors on stock return volatility on the Nairobi Securities Exchange in Kenya. Macro-economic variables used in the study were, foreign exchange rate, interest rate and inflation rate. The study used monthly time series data for ten years between January 2001 and December 2010.The study used EGARCH and TGARCH models to investigate the relationship between economic variables and the stock market return. Findings in the study showed evidence that foreign exchange rate, interest rate and inflation rate affect stock market returns volatility. The study used ADF test to test the unit root and the Jarque bera test to test for normality in the time series data variables. 2.4.6 Relationship between Investor Herding Behavior and Stock Market Volatility Empirical studies show that investor herding behavior leads to inefficiency in stock markets where asset prices display significant deviations from prices expected in an efficient market. According to Gelos and Wei (2002), emerging capital markets constitute environments whose institutional structures naturally facilitate the manifestation of herd behavior. Studies like Blasco et al. (2008), Patterson and Sharma (2006), Chiang et al. (2011), Zafer (2012), Voronkova and Bohol. (2003), Nofsinger et al. (1999) and Seetharam and Britten (2013) have examined herding behavior in various markets and returned diverse findings on its effect on stock prices and market volatility. 41 Blasco , Corredor and Ferreruela (2008) examined the implications of herding on volatility of the Spanish stock market. The study covered the period between January 1997 and December 2003. Intraday data was used to calculate the herding measure. Herding was measured using the Patterson and S