A Study on the Factors affecting the Stock Market Returns in Malaysia

Table of Contents

Chapter 1: Introduction to the Study
1.0. Introduction
1.1. Background of the study
1.2. Problem Statement
1.3. Objectives of the Research
1.3.1. General Objective
1.3.2. Specific Objectives
1.4. Research Questions
1.5. Significance of the Study
1.6. Scope of the Study
1.7. Conclusion

Chapter 2: Literature Review
2.0. Introduction
2.1. Literature Review
2.1.1. E-change Rates
2.1.2. Inflation
2.1.3. Crude Oil Prices
2.1.4. Foreign Direct Investment
2.2. Conclusion

Chapter 3: Research Methodology
3.0. Introduction
3.1. Theoretical Methodology
3.1.1. Research Philosophy
3.1.2. Research Approach
3.1.3. Research Strategy
3.1.4. Time Horizons
3.2. Practical Methodology
3.2.1. Theoretical Framework
3.2.2. Conceptual Framework
3.2.3. Research Design
3.2.4. Data Collection Method
3.2.5. Sampling Design
3.2.6. Data Analysis
3.2.7. Diagnostic Checks
3.2.8. Granger Causality Test
3.2.9. Multiple Regressions Model
3.2.10. Research Hypothesis
3.3. Conclusion

Chapter 4: Data Analysis
4.0. Introduction
4.1. Unit Root Test
4.1.1. Augmented Dickey-Fuller (ADF) Test
4.1.2. Phillips Perron (PP) Test
4.2. Diagnostic Checks
4.2.1. Multicollinearity
4.2.2. Heteroskedasticity
4.2.3. Autocorrelation
4.2.4. Model Specification
4.2.5. Normality Test
4.3. Granger Causality Test
4.4. Regression Model
4.4.1. T-test
4.4.2. F-test
4.5. Conclusion

Chapter 5: Findings, Discussions and Implications
5.0. Introduction
5.1. Summary of Unit Root Test
5.2. Summary of Diagnostic Tests
5.2.1. Multicollinearity
5.2.2. Heteroskedasticity
5.2.3. Autocorrelation
5.2.4. Model Specification
5.2.5. Normality
5.3. Summary of Granger Causality Test
5.4. Summary of OLS Regression
5.5. Discussion on Major Findings
5.6. Implication of the Study
5.7. Conclusion

Chapter 6: Limitations, Summary and Recommendations
6.0. Introduction
6.1. Limitations
6.2. Summary of the Study
6.3. Recommendations for Future Research

7.0. References

Appendices

List of Tables

Table 1: Source of Data

Table 2: Research Hypothesis

Table 3: Results of ADF Test

Table 4: Results of PP Test

Table 5: Correlation Test of Variables

Table 6: Heteroskedasticity Test (ARCH)

Table 7: Breush-Godfrey Serial Correlation LM Test

Table 8: Ramsey RESET test Output

Table 9: Granger Causality Test Results

Table 10: OLS Regression Result of the Model

Table 11: T- test Results

Table 12: F- test Results

Table 13: Summary of Diagnostic Tests

List of Figures

Figure 1: Theoretical Framework

Figure 2: Diagram of Data Processing

Figure 3: Regression Result of lnCPI and lnFDI

Figure 4: Jarque-Bera Normality Test

Figure 5: OLS Regression Results

Acronyms

ADF - Augmented Dickey Fuller APT - Arbitrage Pricing Theory

ARCH - Autoregressive Conditional Heteroskedasticity BLUE - Best Linear Unbiased Estimator

BNM - Bank Negara Malaysia

CAPM - Capital Asset Pricing Model

CLRM - Clear Linear Regression Model CPI - Consumer Price Inde-

CPO - Crude Oil Price DW - Durbin Watson ER - E-change Rate et al. - And others

E-Views - Econometric Views FDI - Foreign Direct Investment FET - Fisher Effect Theory GDP - Gross Domestic Product GFC - Global Financial Crisis IR - Inflation Rate

JB-Statistics - jarque-Bera Statistics

KLCI - Kuala Lumpur Composite Inde- LN - Natural Logarithm

MPT - Modern Portfolio Theory OLS - Ordinary Least Square PP - Phillips-Perron

RESET - Regression Equation Specification Error Term

RM - Ringgit Malaysia

VIF - Variance Inflation Factor

Dedication

This research project is e-clusively dedicated to my parents Filipe Chauque and Isabel Ale-andre Mandlate as an appreciation for their overwhelming support and encouragement throughout the completion of my bachelor degree.

Acknowledgement

First and foremost, my appreciation goes to Albukhary Foundation, for providing me with this opportunity to pursue my higher education in Asia Pacific University. I really appreciate their support.

Secondly, I would like to e-press my deepest thankfulness and appreciation to my supervisor Ms. Patricia A/P Rayappan for her valuable time, support and advices given to me throughout the journey of completing this research. Without her guidance this research would have not been as perfect as it is. Apart from that, I also would like to thank all the Asia Pacific University lectures for sharing their valuable knowledge and opinions with me whenever I encountered any problem.

Last but not least, my appreciation goes to my family, friends and classmates for their assistance in terms of morale and ideas. Their indirect contribution is also gratefully acknowledged.

Abstract

The current research paper investigates the dynamic relationship between Kuala Lumpur Composite Inde- (KLCI) and four selected macroeconomic variables namely e-change rate, inflation rate, crude oil price and foreign direct investment. The research consists of 108 monthly observations from the period of January 2007 to December 2015. In this research, the Augmented Dickey-Fuller test (ADF) showed that at 5% significance level, all the variables are stationary at first difference. For the diagnostic tests, there is no multicollinearity, heteroscedasticity, autocorrelation, and model specification problems. However, normality problem was detected in the model. Moreover, Granger causality test and OLS regression model were carried out to determine the short-run and long-run relationships between the KLCI and the selected macroeconomic variables respectively. Results suggest that in the short-run there is no relationship between the KLCI and the four selected macroeconomic variables. However, in the long-run e-change rate, inflation rate, and crude oil prices are found to significantly affect the performance of KLCI, whereas foreign direct investment is found not to influence the movements of KLCI. The e-change rate and inflation negatively affect the KLCI, and the crude oil price has a positive impact in the KLCI movements.

Chapter 1: Introduction to the Study

1.0. Introduction

Stock market can be defined as a market place, where stocks and shares are the traded commodities, (Masoud, 2013). The importance of the stock market in the development of the economy of a country can be directly linked to the governance, appropriate and effective regulatory framework designed by the policymakers. The Malaysia stock market plays a very significant role in promoting capital formation and sustaining the economic growth of the country. It efficiently allocates scarce resources which are used to finance different sort of projects, leading to the prosperity and growth of the economy. Moreover, a healthy stock market serves as a vehicle for risk diversification associated with projects, thus minimizing the uncertainty regarding returns. The participants in this market are typically large institutional investors such as mutual funds, insurance companies, pension funds, investment banks, hedge funders, as well as individual investors. These investors usually earn money through the stock market by buying and selling shares if they are proficient in predicting the movements of stock market.

According to the Economic Times (2016), volatility is a rate at which the price of securities fluctuates for a particular set of returns. If the stock prices change constantly over a short period of time, the volatility is considered high. On the other hand, if the stock prices are stable, or do not change significantly, the volatility is considered low. High stock market volatility leads to a huge variation of returns, and thus greater risk (Lim & Sek, 2013). Although volatility can bring huge losses to investors, it is through this that investors may also make profits. Therefore, having a good understanding of the volatility of the stock market leads investors to a more precise prediction of the stock price movements, which in turn reduces the risk of making losses and may increase returns. Nevertheless, the risk associated with investing in stock market can also be reduced by having a well-diversified portfolio of assets.

Olweny and Kimani (2011) have argued that stock market facilitates the investment of surplus funds into additional financial instruments that better match their liquidity preference and risk appetite. In this new era of globalization, the unpredictability of the stock market returns has become a major subject in developing countries like Malaysia. Stock market plays an important role as a source for new private capital in the country. It also symbolizes the growth and economic activity of the country. For these reasons, the central bank of Malaysia, Bank Negara Malaysia (BNM) and the government have to pay a close attention in the behavior of the stock market, because an increase or decrease in the stock prices could reflect future growth or recession of the economy respectively. According to Kutty (2010), the significant correlation between stock price and macroeconomic variables is essential to policymakers, economists, and investors. This relationship helps them to better access the efficiency of the market during portfolio management.

The stock market is an essential part of the economy of many developed and developing countries. The Malaysia stock market is an essential tool in facilitating the flow of money between people who want to save and people who want to use the money for investments or other particular reasons. It also establishes the flow of resources to the most profitable investment opportunities. According to Nordin & Nordin (2016), at the end of 2010 the stock market capitalization and debt outstanding stood at 165% and 97% of nominal GDP respectively. This figures show that the Malaysian capital market is quite big relatively to the Malaysian economy. Therefore, taking into account the size of the Malaysian capital market, it is possible that this market could significantly contribute to the economic growth of the country. Moreover, with a stable growth in the last years, the stock market in Malaysia is e-pected to play a major role in the global financial market, which will turn to provide attractive investment opportunities for foreign investors.

Considering the uncertainty associated with the stock market movements, there are several macroeconomic factors that are found to be important in estimating the relationship between the stock market returns and its respective factors in Malaysia. The current research paper suggests that e-change rate (Yusuf and Rahman, 2012), inflation rate (Chinzara, 2011), oil price (Hossenidoust et al. 2013) and foreign direct investment (Ali, 2014), have a significant correlation with the stock market returns in Malaysia.

Economists have often argued about the relationship between the stock market and e-change rates due to their impact in influencing a country’s economic development. According to Yusuf and Rahman (2012), when the stock market is optimistic, foreign investors would come and invest in Malaysia, and the value of the ringgit would appreciate and vice versa. On the other hand, when the ringgit is very volatile, foreign investors would not want to hold the Malaysian stocks due to the fear of foreign e-change risk incurred which would affect the Malaysian equity market. Therefore, e-change rate is assumed to be a major factor in determining the returns of the stock market in Malaysia.

Inflation is usually referred as a rise in the general prices of goods and services. It happens when people spend higher amount of money to buy the same or lesser quantity of products they used to buy before. Controlling inflation has been a priority goal for monetary authorities of many countries, especially those under development. Having inflation rate under control creates an attractive environment for investment in any country. According to Chinzara (2011), inflation rate is a major factor affecting the stock market volatility. In the research on macroeconomic uncertainty and volatility of the stock market, the researcher also found that the movements of stock prices volatility are significantly influenced by several macroeconomic uncertainties. Moreover, a study conducted by Geetha et al (2011), also found that in the long run, inflation rate adversely affect the stock market in US, Malaysia and China. Hence, inflation rate can also be another determinant of stock market returns.

The volatility of oil prices can hugely affect the performance of various industries as well as the economy of a country as a whole. In the last few years, the world has witnessed a tremendous decline in the price of oil. This decline in oil prices has affected many countries, especially Malaysia as an oil e-porter country, contributing in the depreciation of the local currency. Studies conducted before have investigated the correlation between the stock market and oil prices. Arouri et al. (2011), has found that oil prices not only determine but also predict stock returns. Hossenidoust et al. (2013) have also studied the influence of prices of oil on some ASEAN stock markets. They provide evidences of the significant influence of the oil prices on Singapore and Malaysia stock market’s returns. Therefore, oil prices may have a huge influence in the stock market returns not only in Malaysia, but also in some other neighbor countries.

FDI is considered to be one of the essential sources of e-panding investments in countries under development like Malaysia. It involves e-panding operations in a country either by a foreign or local corporation. In the same way, stock market also does boost the amount of investment opportunities contributing to the economic growth of a country. Ali (2014) has studied the impact of FDI on stock market returns. The researcher found that there is a strong positive correlation between stock market and foreign direct investment. Both variables are found to increase the investment opportunities which in turn increase economic development of the country. Accordingly, FDI is considered an essential factor in estimating stock market returns.

Notwithstanding, movements in the stock market can have a deep impact on the economy and consequently on the people’s daily life. A collapse in stock prices can easily cause disruption to the economy of any country. Back to 1930s, the stock market crash of 1929 was the main factor that caused the great depression, which resulted in decline in the economic activity worldwide. In addition, Farmer (2015) argues that the fall of the stock market in 2008, provides an applausive e-planation for the magnitude of the great recession.

1.1. Background of the study

The Malaysia stock market currently known as Bursa Malaysia is the only stock e-change in the country. It was established in 1964 with the aim of providing to governments, companies and other individuals a platform to trade securities, including other services such as depository, clearing, and settlement services. Bursa Malaysia plays an important role in assisting the growth of Malaysia capital market by offering services and infrastructures that enable the creation of a competitive global marketplace through adopting internationally accepted standards (FTSE Bursa Malaysia KLCI, 2011). The Bursa Malaysia is committed to maintain an active, efficient, and secure trading market for both domestic and foreign investors.

The Kuala Lumpur Composite Inde- (KLCI) which is also known as FTSE Bursa Malaysia is the representative of the Malaysian stock market inde- since 1986, comprising the top 30 largest companies in the country. According to Chong and Puah (2009), Kuala Lumpur Stock E-change Composite Inde- (KLCI) is a capitalization weighted inde- which is used as an accurate indicator to measure the Malaysian stock market performance. Comparing the Malaysian stock market and other stock markets around the world, Bursa Malaysia is considerably new and still categorized as an emerging market. However, since its establishment it has been e-panding rapidly, to become one of the major emerging markets in Asia-pacific.

During the years of 2007 and 2008, the Global Financial Crisis (GFC) affected all the financial markets in the world, causing a massive outbreak to the Malaysia economy which contributed in the creation of an enormous fluctuation of the KLCI, which is the main inde- and market indicator in Malaysia. Thus, it is relevant to e-amine the Malaysia stock market performance using significant macroeconomic variables such as those that played a major role in the last global financial crisis.

Majority of scholars have agreed that the fluctuation of the stock market returns is affected by different types of interrelated economic factors. For instance, foreign investment in a country can increase if the stock prices are in an upward trend. On the other hand, a drop in stock price will negatively affect corporate returns which in turn will reduce the country’s wealth. Consequently, due a drop in the stock price, currency is more likely to depreciate. Thus, stock market has become crucial in promoting capital formation and sustaining economic growth in a country.

Studies conducted before agree that the Malaysian stock market is very sensitive to the economic conditions of the country. Therefore, investors could use the country’s level of economic activities to determine or forecast the future performance of companies. When the economy is performing badly, the growth of companies is compromised leading the stock price to drop. On the other hand, when the economy is performing pretty well, the level of production increases, and there is a significant decrease in the unemployment level, which in turn influence the stock price to rise. Hence, stock price can be used as an important determinant of the economic activities as well as equity market return, since the stock market return is always affected by changes in the economic conditions. Hereafter, the current paper focus on investigating whether macroeconomic variables such as e-change rate, oil prices, inflation rate, and foreign direct investment affect the Malaysian stock market returns.

1.2. Problem Statement

Among macroeconomists and finance theorists there is a mutual consensus that stock market returns are driven by macroeconomic factors which are considered as essential in the economy. The Black Monday on 19th October 1987, the Wall Street Crash of 1929, the Asian Financial Crisis (AFC) of 1997 and the recent GFC of 2008 are applausive e-amples of how the stock market can affect the domestic and worldwide economies.

Although the performance of Bursa Malaysia was badly affected by changes in macroeconomic variables in 1997 during the AFC and 2008 during the GFC, few studies have been conducted in relation to these macroeconomic factors impact on the stock market. Before 1996, KLCI reached more than 1200 points and investors e-pected the trend to continue to sustain. However, the Inde- dropped to 500 points due to the 1997 AFC (Asmy, Rohinila, Hassama, & Fouad, 2009). Moreover, according to Angabini & Wasiuzzaman (2011), the KLCI also fell drastically around 558.93 points in 2008 which was equivalent to a drop of around 40% in its value due to the GFC. This drop in value of KLCI caused huge losses for stock market investors, and a large number of them accumulated some huge degree of financial debt.

Since the GFC of 2008, the KLCI recorded its first decline in value in the end of 2014, although in smaller magnitude. According to Amanah Mutual Berhad (2015) the KLCI dropped by 5.7% in value, a smaller scale compared to the 40% registered last time in 2008. This decline in value was mainly caused by the e-tended decrease in the prices of crude oil which was previously being traded above $100 per barrel in the beginning of 2014, and below $50 in the end of the same year.

In the view of these declines in the value of KLCI inde-, investors will always be interested in tracking the movement of the stock market in order to be able to make rational investment decisions. Therefore, it is important to e-plore the relationship between the Malaysian stock market and macroeconomic variables, to shed light on the relationship, if there is any, between the economic activities and the stock market performance in Malaysia.

Several studies have been conducted in advanced economies such as the U.S, United Kingdom (UK), Japan and Germany, in the impact of macroeconomic variables on stock market returns. However, the results from these studies cannot be generalized and assumed to have the same impact in Malaysian market since the country is still under development, and the stock market is relatively new compared to other advanced economies. Poole (2010) has stated that variations in the macroeconomic conditions of a country could help to forecast variations on the stock market. Nevertheless, the researcher was not clear in how macroeconomic conditions impact the stock market returns. Consequently, in the light of these gaps, and also the e-istence of few studies conducted in emerging economies, the current research paper aims to e-amine the dynamic relationship between these macroeconomic variables (e-change rate, inflation rate, oil prices and FDI) and stock market returns in Malaysia.

The importance of e-ploring this relationship is of e-treme importance for investors given that the key risks they face in the stock market might be traced back to changes in the macroeconomic factors. Understanding the dynamic behavior of the stock market is essential not only for investors, but also for financial analysts, policymakers and macroeconomists. Furthermore, this relationship helps financial analysts and investors to develop accurate models which may help in analyzing the risk of holding financial instruments. Thus, the objective is to observe whether macroeconomic elements individually or collectively play a role in the Malaysian stock market returns.

1.3. Objectives of the Research

1.3.1. General Objective

The general objective of this paper is to e-amine the dynamic relationship between stock market returns and macroeconomic variables in Malaysia.

1.3.2. Specific Objectives

- To investigate the relationship between e-change rate and the stock market returns. -To investigate the correlation between inflation rate and the stock market returns. -To investigate the correlation between oil prices and the stock market returns. -To investigate the relationship between FDI and the stock market returns.

1.4. Research Questions

- How does the fluctuation of e-change rate influence the stock market returns?
- Is there any significant correlation between stock market returns and inflation rate? -Does foreign direct investment influence the stock market returns?
- How does the volatility of oil prices influence the stock market returns?

1.5. Significance of the Study

The current paper aims to investigate the influence that macroeconomic variables such as e-change rates, inflation rate, foreign direct investment and oil prices may have on the stock market returns in Malaysia. The findings of this research will provide a lot of meaningful information that will benefit a wide range of people such as policymakers, personal and corporate investors, portfolio managers, hedgers, speculators, and academicians. The findings from this paper are e-pected to assist the stock market participants in the diversification of their portfolio in order to reduce the risk associated with stock market investment. Hence, this research paper will enhance the significance of risk management, investment decisions and also address the influence that macroeconomic variables have in e-plaining Malaysian stock market returns. Since major movements in macroeconomic activities can influence the fluctuation of the stock market returns, the findings in this paper will help investors to understand how the stock market responds to macroeconomic variables as non-diversifiable components.

From the investors’ perspective, individual investors, portfolio managers and fund managers may take a look into the findings of the current research to decide on whether to buy or sell stock. The paper also provides to investors a better understanding on whether and how stock market returns respond to the volatility of macroeconomic variables. Based on this research, managers will be able to allocate their resources properly in order to increase the returns by ma-imizing the stock prices. This research will also help managers to hedge the stock by formulating better investment decisions, diversifying their investment portfolios, change investment strategies which in turn will minimize the risk of losses.

From the government’s perspective, understanding the correlation between the returns coming from the stock market and macroeconomic variables can be a useful tool for policymakers to predict future challenges in the economy as well as in the stock market. Therefore, findings from this paper would help regulatory bodies to implement appropriate macroeconomic policies that can stabilize the financial market and avoid the volatility of the stock market returns. By doing so, the government would be enhancing the stock market industry, reducing the uncertainty regarding stock market returns, which would attract more investors either locally or internationally to invest in the Malaysian stock market due to favorable investment environment in the country.

Last but not least, academicians can also use this research paper as a tool to develop new hypothesis that can enhance future studies on stock market.

1.6. Scope of the Study

This paper starts with an introduction of the topic to be studied. The introduction comes with a basic e-planation of the topic and its relevancy and also a background of the study. Then it is followed by the objectives of the research, research questions, significance of the study, and the layout of the study. The second section provides some basic concepts and review of the of the previous literatures in the impact of macroeconomic variables on stock market returns of different countries. The section three is regarding the data collection and the methodology to be used in the research to test the hypotheses. The following section analysis the data collected from the secondary sources. The section five summarizes the findings, and provides discussion on the major findings derived from the data analysis, as well as its implications. The last chapter of the study points out the limitations of the research, followed by a summary of the first five chapters and also provides recommendations for future researches.

1.7. Conclusion

In the nutshell, this chapter discusses the importance of investigating the macroeconomic variables impact on stock market returns in Malaysia. The chapter e-plains the Malaysia stock market’s background in order to provide a perfect picture and a better understanding about the Malaysia stock market. The chapter discuss about the problem statement, research objective, research question, and significant variables for the study. It also e-plains the relevance of analyzing the dynamic relationship between the macroeconomic variables (e-change rates, inflation rates, oil prices and foreign direct investment) and stock market returns.

Chapter 2: Literature Review

2.0. Introduction

The current chapter conducts a review of the literature on stock market returns and macroeconomic variables. The review of the literature will be based on the work done by other researchers on the impact of e-change rates, inflation, oil prices and foreign direct investment on stock market returns. Specifically, this chapter provides the theoretical framework guiding this research paper, and it also tries to e-plain the correlation between stock market returns and macroeconomic variables in Malaysia from the empirical results obtained.

2.1. Literature Review

Stock market plays a major role as a source of new private capital in the country by bringing together businesses and entrepreneurs to trade stocks for the purpose of capitalizing enterprises that need cash. It also symbolizes the growth and economic activity of many countries, including countries under development like Malaysia. Therefore, many researchers have been interested in identifying and e-amining the factors that affect the stock market due to its importance in the economy.

A lot of studies have been conducted in order to determine the factors driving the stock market performance using different macroeconomic variables. For instance, Mohanty et al. (2011), have used the volatility of oil price to determine its impact on stock market at different country and industry level. Moreover, Hsing (2011), used variable such as e-change rate to test the stock market inde- in some European countries. Raze et al. (2012) e-amined the effect that foreign direct investment has on the performance of stock market in Pakistan. The same study was also conducted by Ali et al. (2010), however the main variable in this research was the inflation rate.

According to Osho (2014), stock market being a major element in the financial sector of several developing countries, it serves as a fundamental element towards economic development through diversification, mobilizing funds, and pooling together funds of people who want to save and people who want to borrow, for its optimal utilization. Stock market serves as a source of capital for both government and private institutions. However, it also offers the opportunity to diversify the financial sector services, and allow investors to put their money in a safer place, depending on how stable the market is. Stock market is a pivotal element in developing a strong and competitive economy.

2.1.1. E-change Rates

The relationship between the e-change rate and stock market returns has been a subject of study by many researchers since both are considered as crucial elements in influencing the economic development of many countries. Though, the findings regarding their influence on the stock market returns appear to be very controversial.

According to Kutty (2010) being able to understand the relationship between the e-change rate and stock market returns is an important factor from the point of view of policy makers and investors in this changing global environment. Movements of the e-change rate can have a huge impact on the cash flows of multinational companies, since the performance of these companies not only depend on the resources that companies possess, but also on the fluctuations of the currencies, assuming there will always be a conversion from one currency to another.

Cakan & Ejara (2013) who studied the relationship between the e-change rates and stock prices, have argued that when the local currency of a country depreciates, it makes e-porting goods much cheaper and this can lead to a rise in foreign demand and sales. On the other hand, when the local currency appreciates foreign demand of an e-porting firm's products shrinks so the profit of the companies will decrease as well as its stock price. The study has investigated the dynamic relationship between the two variables for twelve emerging market countries from May 1994 to April 2010 using linear and non-linear Granger causality tests. Findings from the tests suggest that majority of the countries studied have both linear and non-linear bi-directional causality in most of the cases, e-cept for three countries where there is no evidence for a nonlinear Granger causality from e-change rate to stock prices.

Agrawal et al. (2010) conducted a study on the correlation between the Nifty returns and India rupee-US Dollar e-change rate using daily closing inde- from October 2007, to March 2009. Findings from this study revealed that there is a negative relationship between the Nifty returns and e-change rate during the period analyzed. Similarly, Najaf & Najaf (2016) used Granger causality test to check the level causal relationship between the two variables in the Indian market. The objective of the study was to prove that e-change rate is a crucial determinant of firms’ profitability. However, results show that movements in the e-change rates negatively affect the stock prices. From these arguments we may e-pect a causal effect of e-change rates on stock market returns in India.

Younas et al. (2013) also found a negative correlation between e-change rate and stock price in his study in the impact of e-change rate on stock market in Pakistan. For this purpose, they used a Generalized ARCH model and the e-tension of Autoregressive Conditional Heteroscedasticity (ARCH) to conduct the research, using daily time series data for the period of 2002 to 2012. This study suggests that e-change rate not only affect the returns of multinational companies, but also affect the returns of domestic firms. From multinational companies’ perspective, e-change rates bring a sudden change in worth of its foreign operations which in turn reduces profitability and affect the stock price negatively. While in the case of import oriented domestic firms, stock prices will negatively be influenced due to the currency depreciation which leads to an increase in the price of inputs which results in a decrease of the profitability of the firms.

Milambo et al. (2013) who studied the impact of e-change rate volatility on South Africa stock market used the GARCH model to establish the relationship between the two variables employing monthly data from the period of 2000 to 2010. Findings from this study suggest that movements of the currency have a huge influence on the value of the rand of cash flows from foreign projects. However, the study also suggests a very weak correlation between the volatility of the currency and the stock market, but a huge impact in the financial system. In addition, it was found that the South African stock market is affected by other macroeconomics variables such as total mining production, interest rates, money supply, and the United States interest rates.

Muhamad and Rasheed (2011) have e-amined the correlation of stock prices and e-change rates in four South Asian countries namely Pakistan, India, Bangladesh and Sri-Lanka for the period of January 1994 to December 2000. The main objective was to e-amine the long-run and short association between the variables in the four countries. The results provide evidences that for Pakistan and India there is neither long-run nor short-run association between the e-change rates and the stock market prices. Moreover, no short-run relationship was found for Bangladesh and Sri-Lanka. Notwithstanding, the results appear to suggest a bi-directional long-run causality between the two variables in the case of Bangladesh and Sri-Lanka. The results also reveal that in the short-run, stock prices and e-change rates are not related in South Asian Countries. For that reason, information from one market cannot be used to forecast the other market.

Hypothesis 1: Impact of e-change rates on Malaysian stock market returns.

2.1.2. Inflation

Inflation is considered as a general rise in the price of goods and services in the economy. Inflation is undoubtedly one of the most important macroeconomic variables believed to be related to stock prices, and in turn also affected by it (Gupta and Inglesi-Lotz, 2012).

High inflation brings a lot of problems to the economy of a country, since it reduces the volume of output, and thus companies tend to lay off workers, which in turn increase the level of unemployment in the country. Therefore, the government by its Central Bank or monetary authorities has to implement policies to contain the inflation, usually by increasing the interest rates which reduces the amount of money circulating in the market.

According to Fatimah & Shamim (2012) an increase in the interest rates will cause a drop in the stock price since both have an inverse relationship. Interest rates and inflation are two highly correlated variables, since one variable drives another up and down, depending on the economic condition of the country. This argument suggests that the stock prices can also be affected by inflation since both variables are influenced by the level of interest rate.

Adusel (2014) investigated the correlation between the inflation and stock market returns from Ghana Stock E-change for the period of January 1992 to December 2010. The study found that there is a negative statistical significant relationship between inflation and stock market returns in the short run. However, this negative relationship becomes significantly positive in the long run. The negative short-run between the inflation and stock market returns suggests that a rise in the inflation will cause a drop in the price of stock market. Although, in long-run inflation pushes the stock market returns towards equilibrium, even though the speed of adjustment is very slow. The researcher used Cointegration and Granger Causality test in the Error Correction Model to analyze and also employed autoregressive distributed lag (ARDL) bounds testing approach cointegration to find whether there is long run between inflation and stock market returns in Ghana.

Mousa et al. (2012) used time series data from the Consumer Price Inde- (CPI) as a measure of inflation and the stock prices of ten selected companies in Jordan as a measure of stock validation to test whether there is a correlation between stock prices and inflation. Findings from the study suggest that majority of the companies e-amined (70%) are negatively correlated against inflation, whereas the rest (30%) show a slightly positive relationship between changes in the stock prices and inflation. Moreover, results show that stocks cannot be used as a perfect hedge to the degree that firms cash flow are negatively correlated to inflation, and the relationship between stock price and inflation can be either negative or positive.

Nader & Alraimony (2012) studied the impact that macroeconomic factors have on stock market returns in Jordan. They used monthly data instead of quarterly for the period of 1991 to 2010 to have larger number of observations and capture the long term movements by using ARCH/GARCH model. Findings also proved that there is an inverse relationship between the stock prices in Jordan and inflation, where the coefficient of correlation was highly significant.

Geetha et al. (2011) analyzed the impact of inflation on stock market of three countries namely: Malaysia, United States and China. The aim of the study was to identify whether there is a correlation between inflation and stock returns based on the level of development of the three countries’ economy. The researchers used secondary data consisting of monthly time series data from January 2000 to November 2009. They also used interest rate, inflation (CPI), e-change rate, GDP, and share prices of the three countries as variables of the study. The study found that there is long run correlation between inflation either e-pected or une-pected with stock returns, but there is no short run correlation between these variables for Malaysia and US, but it e-ists for China.

Uwubanmwen & Eghosa (2015) conducted a research in the impact that inflation rate have on stock returns in the Nigeria stock market. The study also tried to determine whether stock returns in Nigerian stock market were influenced by the inflation rate and also to establish whether stock returns in the Nigerian stock market can effectively be forecasted using stock prices. For this purpose, the researchers analyzed monthly data for the period of 1995 to 2010. The data was e-tracted from the Nigeria Stock E-change Fact Book and the Central Bank of Nigeria Statistical Bulletin. In order to estimate the relationship between inflation rate and stock returns the method used was autoregressive distributed lags (ARD). Findings suggest that there is a negative but weak influence of inflation on stock returns. Therefore, inflation cannot be used as a crucial element to predict the stock returns in Nigeria.

Ahmed, Islam and Khan, (2016) used Johansen test to investigate the association between inflation and stock returns in Bangladesh. The study used stock return data from monthly closing stock price indices of Dhaka Stock E-change (DSE), and monthly data of inflation rate for the period of November 2004 to July 2013. The Johansen test procedure established the e-istence of a single cointegration equation at 5 percent significance level, which suggests a long run equilibrium correlation between the stock market and inflation. The study also found a short run positive relationship between the stock market and inflation in Bangladesh. The implication of this short run relationship is that une-pected inflation can raise the equity value of firms if they are net debtors. Conversely, if the Central Bank implements monetary policies, inflation can drop and stock prices would rise. The study also hypothesized that stock market returns can adversely be influenced by inflation since inflationary pressure would threaten future profits of companies.

Hypothesis 2: impact of inflation rate on Malaysian stock market returns.

2.1.3. Crude Oil Prices

In the recent months the price of crude oil has dropped drastically. This drop in the price of oil has a significant impact on many sectors of the industry, since the transportation costs are also affected and this effect is reflected in the selling price of goods and services. Zhang & Chen (2011) have argued that crude oil which is considered as one of the world’s most influential commodities plays a very important role in many economies. A lot of researchers have also pointed out that oil shocks can also cause economic recessions.

Reboredo et al. (2013) indicate that changes in the oil prices have no effect in stock market returns in the pre-crisis period at either the aggregate or sectorial level with correlations in all time scales close to zero, e-cluding oil and gas company stocks which were positively influenced the movements of oil price. In addition, the study found that contagion and positive interdependence between oil and stock prices have been identified in Europe and the USA since the beginning of the global financial crisis. The main objective of the study was to analyze the relationship between the oil and the stock market at aggregate and sectorial level, and also contagion and interdependence level between the variables. For this purpose they used wavelet multi-resolution analysis. The study was conducted using daily data for the period of June 2000 to July 2011.

Jafarian & Safari (2015) performed Granger causality test in order to e-amine the causal relationship between oil price and Malaysian stock market returns in different sectors. A multifactor regression model was used to measure the impact that changes in oil prices may have on the return of stock market inde- as well as eight economic sector indices. The study was conducted using monthly data of stock prices of Bursa Malaysia and eight economic sectors of the country. The regression analysis results indicate that the overall stock market represented by KLCI is influenced by changes in the price of oil. However, Granger causality test failed to support any causal relationship. This suggests that Malaysian stock market return is highly dependent on the price of oil, which might be the main source of energy for the whole country. In addition, findings indicate that oil and gas firms can be influenced positively by changes in the price of oil. The higher is the price of oil, the more are the earnings for oil and gas companies. This relationship is proven by performing Granger causality test.

According to Ponka (2016) real oil prices are very useful forecasters of the trend of stock returns in a variety of markets, even after controlling for the prognostic power of commonly used predictors. However, the study also found that overall sign of predictability of returns and predictive power of oil price changes vary significantly between countries. The increase and decrease in the price of oil was found to affect the direction of stock returns, but in some markets there are evidences of some possible asymmetry. In this study, the researcher main objective was to e-amine the predictive power caused by changes in real oil prices on the sign of e-cess in the U.S and 10 other markets using probit models.

Fang & You (2014) studied the effect of oil price shock on big emerging countries’ stock prices namely China, India and Russia. The main objective of the study was to investigate how e-plicit structural shocks that characterize the endogenous character of changes in oil prices affect three large Newly Industrialized Economies (NIE’s) stock-market returns. To conduct this study, the researchers applied monthly data for the period of 2001 to 2012 using global oil production, global real activity data, real oil price, and real stock inde- returns. The researchers implemented the SVAR estimation and IRA in their analysis. Based on the paper’s empirical results, as long as the change in oil price is not driven by an increase in the oil consumption in the case of India, the oil prices always have a negative effect on India’s economy. Secondly, the Russia’s stock returns are positively affected by oil price movements, but only if these movements are driven by Russian oil-specific supply shocks. And finally in the case of China, oil price is driven by global demand shocks has an insignificant effect on China’s financial market.

Tsai (2015) conducted a study to investigate how the United States stock returns respond differently to oil price shocks prior to, during, and after a financial crisis. The researcher used a long time series daily data for 682 firms for the period of January 1990 to December 2012. Each firm includes 5,772 daily observations of oil prices and stock returns. Returns were measured using closing price on day t to the closing price on day t-1. Results provide evidences that increase in the oil prices in the pre-crisis period both statistically and negatively affect the U.S stock returns. On the other hand, empirical results indicate that on the recovery of financial crisis, the e-citing effect on stock returns through a rise in oil prices dominates the negative effect of the rise in production cost. Therefore, the U.S stock returns positively respond to the changes of oil prices during both the crisis and after the crisis.

Mollick & Assefa (2013) used GARCH and MGARCH-DCC models to establish the stock returns and oil price relationship in the United States. The study is conducted using daily data for the period of January 1999 to December 2011. The researchers used the U.S stock price indices to calculate the daily returns namely: S&P 500 COMPOSITE, Dow Jones Industrial Average, NASDAQ, and Russell 2000. The study suggests that prior to the financial crisis the stock returns are slightly negatively affected by oil prices. However, during the aftermath of the crisis which includes higher demand of oil together with more e-ports to the rest of the world, the U.S stocks respond positively to e-pectations of recovery.

Hypothesis 3: Impact of oil prices on Malaysian stock market returns.

2.1.4. Foreign Direct Investment

With the lack of domestic investment which is an important source for the economic development of an economy, FDI plays a crucial role as the main source of international finance cash inflow. According to Melnyk (2014), when FDI comes to a native country, firms receive competitive advantage due to the usage of new knowledge, e-perience, ways of production and management. Yu et al. (2011), consider FDI as one of the most significant channels of technological transfer. A lot of studies have been conducted in different countries in order to find the relationship between the stock returns and FDI.

Ali (2014) e-amined the effect of FDI on volatility of stock market in Pakistan. The researcher found a strong relationship between the FDI and stock market. This relationship can be understood by comparing their roles in the economy: FDI increases the economic development of a country and also supports the stock market development. Similarly, stock market has a positive effect on the economy, and it also increases the investment opportunities in the country. Therefore, both variables are necessary for the growth of an economy. The study used market capitalization functioning FDI, GNP, IFL to measure the stock market, and the FDI and stock market relationship is established by performing the Co-integration test. Similarly, Shahbaz et al. (2013) also conducted a study to investigate the effect of FDI on stock market growth in Pakistan. Findings from the research also indicate a significant positive correlation between the FDI and stock market growth in Pakistan.

Arcabic et al. (2013) used both Engle-Granger and Johansen Co-integration approach to e-amine the e-istence and characteristics of both the long run and short run relationship between the FDI and stock market return in Croatia. According to the study, in the short run FDI positively affects the stock market in Croatia. Moreover, the research indicates that if FDI stimulates rapid technological development and economic growth through the transfer of know-how and technology, then in the long run FDI indirectly affects the stock market growth as well. Moreover, the presence of FDI inflows encourages policymakers to implement market-friendly regulations, increasing the confidence of the investors. When the number of investors increases, it stimulates development and volume of trade on the domestic stock market.

According to Acheapong & Wiafe (2013) FDI has a positive effect on stock market development. This result is similar to the study conducted by Adam & Tweneboah (2009) which provides evidences that FDI influences stock market growth in Ghana. In this study as well, the researchers conducted an investigation on the influence of FDI on stock market development in Ghana. For this purpose, they used quarterly time series data on the variables defined covering the period 1990 to 2010 to undertake the estimation and analysis. The data was collected from International Financial Statistics, Bank of Ghana and International Statistics Yearbook, published by International Monetary Fund.

Olugbenga & Grace (2015) studied the impact of FDI on Nigeria capital market development. The fundamental objective of the research was to know whether FDI influence the Nigerian stock market development considering its role of later in stimulating the growth of the economy of the country. The study employed ADF unit root test and Johansen co-integration test to analyze the data. The data was collected from secondary sources such as publications by the Central Bank of Nigeria, Economic and Financial bulletin, Nigerian Stock E-change fact books for the period of 1970 to 2010. In order to measure the effect of foreign private investment and capital market development in Nigeria, the researchers used Ordinary Least Square (OLS) regression method. The results of the analysis indicate that in the short run the relationship between the capital market growth and FDIs is positive and insignificant. This result is similar to the study of Baghebo and Edoumiekumo (2012), although they used different variables.

Musa & Ibrahim (2014) employed the Johansen co-integration and the error correction mechanism (ECM) techniques to e-amine the impact of FDI and macroeconomic stability on the level of development of the Nigerian stock market over the period of 1981 to 2010. This study revealed that there is a relationship between the variables in the long run, and the FDI was found to have a positive but insignificant effect in the stock market development. Moreover, this paper recommends policies that would encourage foreign firms to be listed which in turn would attract more FDI, leading to stock market development.

Hypothesis 4: Impact of FDI on Malaysian stock market returns.

2.2. Conclusion

This chapter covered the review of the literature relating to the relationship between the stock market returns and micro-economic variables based on previous studies. The researchers conducted a variety of different tests to prove the e-istence of a dynamic correlation between the selected variables. Different researchers found different literature gaps between the dependent and independent variables. Moreover, the researchers e-plained the challenges and limitations faced in order to conduct this research.

Chapter 3: Research Methodology

3.0. Introduction

The chapter three gives an e-planation about the theoretical and practical methodology used to conduct this research, as well as hypotheses for both dependent and independent variables. The main objective of the research is to e-amine the impact of macroeconomic variables on stock market returns in Malaysia. In order to achieve this goal, statistical tests and diagnostic checking will be applied to test the research model. Besides the theoretical methodology, this chapter lays out the theoretical framework, the research design, sampling design which includes the target population, the data analysis methods which include testing of stationary of the data, diagnostic checking and development of research analysis.

3.1. Theoretical Methodology

The theoretical methodology provides reasons for choosing the type of research that will be conducted. Therefore, this part e-plains the nature of the research and justification for choosing the tests that will be conducted to find out the relationship between the dependent variable and independent variables.

3.1.1. Research Philosophy

According to Saunders et al. (2012), research philosophy depends on the way how the researcher looks into a particular problem. Therefore, the researcher should choose a research philosophy that helps in answering the research questions. In this stage, researchers decide how they will e-amine the problem.

- Objectivism: portrays the position that social entities e-ist in reality e-ternal to social actors concerned with their e-istence (Saunders et al., 2012). The current research investigates the impact of macroeconomic variables on stock market returns. In reality, it is not possible that stock market returns are influenced by social actors. The stock market is an objective entity. There are rules and regulations that have to be followed, and the performance of the market is influenced by factors such as liquidity and volatility independently of human consciousness.
- Positivism: in this philosophy, the researcher is limited to data collection and interpretation, and usually the research results are observable and quantifiable. Positivism researchers develop hypotheses from the e-isting theory and collect data to test and confirm that such hypotheses can make further development of the theory.

3.1.2. Research Approach

- Deductive: there are approaches used to conduct business research. These approaches depend on whether the researcher will develop hypotheses from e-isting theories and then test such hypotheses (deductive) or the researcher will collect the data first and develop theories after the data has been analyzed (inductive). Deductive approach involves development of theory or hypotheses and after collecting the data, use research methodologies to test the hypotheses rigorously.

Capital market has caught attention of several scholars around the world. Regarding the stock market, a lot of theories have been developed by different researchers, and this research will develop the hypotheses from these e-isting theories. The data will be based on Malaysian market, and through econometric analysis the results will accept or reject the hypotheses.

3.1.3. Research Strategy

- E-planatory study: a study is considered as e-planatory when the researcher tries to establish a causal relationship between the variables. Since in this research we are going to collect data in order to find any dynamic relationship between the variables, we are certainly conducting an e-planatory research.

3.1.4. Time Horizons

- Cross sectional: cross sectional studies usually involve the use of cross-sectional regression, in order to sort out the e-istence and magnitude of causal impact of one or more independent variables upon a dependent variable of interest at a given period of time. Considering that this research is conducted within a given period of time and sorts out whether there is any impact of independent variables on the dependent variables, this study is considered cross sectional.

3.2. Practical Methodology

The practical methodology lays down all the tests and analysis that the researcher will conduct which are relevant in e-plaining the dynamic relationship between dependent and independent variables.

3.2.1. Theoretical Framework

Abbildung in dieser Leseprobe nicht enthalten

The above figure illustrates the proposed framework based on relevant literatures reviewed in this research. It provides a clear relationship between the dependent variable and the independent variables. Therefore, the framework focuses on stock market which is the dependent variable and e-change rate, inflation which will be represented by the Consumer Price Inde- (CPI), crude oil price and foreign direct investment which are the independent variables.

3.2.2. Conceptual Framework

Conceptual framework gives a brief overview of some of the most used theories in e-plaining the stock market returns based on macroeconomic variables.

3.2.2.1. Markowitz Portfolio Theory (MPT)

One of the most important and prominent economic theories in investment and finance is known as MPT developed by Harry Markowitz in 1952. The theory provides an understanding how investors can ma-imize return and reduce risk. MPT is based on the idea that in selecting a portfolio, investors can ma-imize e-pected returns based on a particular market risk, giving emphasis that high risk is an integral part of high returns. The model assumes that when selecting a portfolio, risk-averse investors can only worry about two elements which are the mean and the variance of their investment return. Thus, if investors are able to minimize the risk of a portfolio as measured by the variance of the stock prices, the stock portfolio can be optimized. According to Azizan & Sorooshian (2014) by using Markowitz model the portfolio variance can be minimized by investing in weak or negative securities correlation in the portfolio.

Furthermore, Azizan & Sorooshian (2014) have argued that Markowitz theory suggests that investors should limit their choices of a portfolio to those which are located along the efficient frontier. The efficient frontier is a set of optimal portfolios that offers the ma-imum e-pected return for a defined level of risk. Markowitz proved that for a defined level of risk, an investor can identify certain combinations of assets that ma-imize the e-pected returns. This suggests that by only assessing the risk and e-pected returns of an asset may not be enough to achieve an optimal return, but investing in more than one asset enables investors to construct an efficient portfolio reaping the highest benefits of diversification. MPT emphasizes not putting all the eggs in the same basket; it attempts to quantify the number of eggs that should be put in each of the various specific baskets.

3.2.2.2. Capital Asset Pricing Model (CAPM)

According to Zabarankin et al. (2013) CAPM is one of the most fundamental and influential concepts in modern finance. CAPM is a single factor model developed by Sharpe (1964) and Lintner (1965) based on the work of Markowitz (1952) regarding the portfolio optimization. The model assumes that the e-pected return in a risky security is a linear function correlated with its beta, and this beta is a sufficient measure of risk related to the average returns of a non- diversifiable asset. One of the strengths of CAPM is the ability to identify the dynamic relationship between the risk and return as well as the likelihood of getting profit or loss from assets.

CAPM is one of the most popular models used to forecast stock returns. However, its unrealistic assumptions in the real world have caused some inconsistency in the theory. CAPM assumes that all investors are risk averse, risk taker and they can lend or borrow at the same risk free rate; besides that, all investors have access to perfect and costless information, and there are perfect capital markets, with no ta-es, transaction or regulation cost (Merton, 2010). Due to these assumptions, a lot of studies have argued that CAPM is not able to e-plain the e-pected returns if for instance others factors such earning per ratio, inflation or interest rates are included. Furthermore, this model can only measure the risk of one single period. Therefore, in order to overcome these weaknesses, Arbitrage Pricing Theory (APT) was introduced.

3.2.2.3. Arbitrage Pricing Model (APT)

According to Eita (2012) APT was introduced as an alternative to CAPM to overcome its unrealistic assumptions and empirical shortcomings. APT is a multiple-factor model developed by Ross in 1976. According to Koch (1996) APT provides a detailed e-planation of the nature and origins of the e-istent risk when investing in financial assets, and how investors are usually rewarded by the capital market for bearing risk. This theory assumes that macroeconomic variables have a systematic consequence on stock market returns. According to Maysami and Koh (2000), economic variables will have impact on discount rates, the ability of firms to increase cash flow and future payouts of dividend. Consequently, macroeconomic factors will become risk factors to take into account in equity market. The greater the interest rates, the greater the discount factor and hereafter caused stock prices lower. Thus, APT uses the e-pected return of risky asset and risk premium of macroeconomic variable.

APT is mainly based on the assumption that there are some few important macro-economic elements affecting the returns of an asset. According to Roll and Ross (1976) if we are able to identify these factors, we can gain an intuitive appreciation of the influence they have on portfolio returns.

3.2.2.4. Fisher Effect Theory (FET)

Inflation refers to the increase of the price level of all commodities. Typically, high inflation rate reduces the volume of output as well as the demand and increase the unemployment level. However, there is a relationship between inflation and interest rates. Normally, high inflation and interest rate may cause serious economy problems. Therefore, the government and regulators have to develop policies to control this issue. The FET states that a percentage increase in the degree of inflation rate will cause a percentage increase in the nominal interest rate. Thus, controlling the inflation rate may result in a stable interest rate which would help to sustain the stability of the economy. Fatimah & Shamin (2012) argued that when the interest rate increases, the stock price decline since they have an inverse relationship.

3.2.3. Research Design

In a quantitative research method, the researcher is focused on quantification of collection and data analysis, whereas in qualitative research, the researcher essentially focuses on words and generation of theories (Bryman &Bell, 2007). This paper adopted quantitative design method to conduct the research. Quantitative design approach uses quantitative data, which is any data in numerical or mathematical form such as percentage, inde-, and descriptive statistics which enables the researcher to do the hypothesis testing, measure and analyze the data in arithmetical form. Moreover, the research applies a series of statistical techniques and sample data to measure the effect of macroeconomic variables on stock market returns.

3.2.4. Data Collection Method

3.2.4.1. Secondary Data

The data collected to conduct this research is obtained from secondary sources. The research opted to use secondary data because the sources of the data contain information which is fundamental in answering the research question, building hypotheses and e-plaining the relationship between stock market and macroeconomic elements.

The secondary data used in the study consists of monthly time series data of 9 years from the period of January 2007 to December 2015. The data will be collected from Thomson DataStream.

3.2.4.1.1. Stock Market Return (Inde-)

In line with other studies (Adusel, 2015), the Kuala Lumpur Composite Inde- (KLCI) is used as pro-y to measure the stock market returns in Malaysia. KLCI is measured in inde- form.

3.2.4.1.2. E-change Rates (Inde-)

According Milambo et al. (2013), e-change rate is used as pro-y variable to measure its impact on stock market returns. E-change rate is a quantitative data which the unit of measurement is in inde- number.

3.2.4.1.3. Inflation Rate (Inde-)

El-Nader & Alraimony (2012) used Consumer Price Inde- (CPI) as pro-y to measure the inflation rate. CPI is measured in inde- number.

3.2.4.1.4. Crude Oil Price (RM)

Ponka (2016) used Brent crude oil price as pro-y variable to measure the impact of oil prices on stock market returns. Crude oil price is a quantitative data which the unit of measurement is in Malaysia Ringgit (RM).

3.2.4.1.5. Foreign Direct Investment (RM).

According to Ali (2014), foreign direct investment is measured by amount of foreign investment. Foreign direct investment is a quantitative data which the unit of measurement is in Malaysia Ringgit (RM).

Table 1: Source of Data

Abbildung in dieser Leseprobe nicht enthalten

3.2.5. Sampling Design

3.2.5.1. Target Population

The target population of this study is on Malaysia Financial Stock Market (KLSE). The study aims to estimate the dynamic relationship between macroeconomic variables and stock market returns in Malaysia. Therefore, we will analyze the Kuala Lumpur Composite Inde- (KLCI) which is recognized as a stock market inde- since 1986 and represents the top 30 companies in the country. Bursa Malaysia which is now known as the FTSE Bursa Malaysia KLCI acts as an accurate performance indicator of the Malaysian stock market as well as the worldwide economy. Thus, our focus is to e-amine the KLCI in order to find the relationship between its performance and macroeconomic variables. In addition, the research will include 108 observations for each variable, which will consist of monthly data collected from the period of January 2007 to December 2015.

Figure 2: Diagram of Data Processing

Abbildung in dieser Leseprobe nicht enthalten

The data processing is based on four steps. Firstly, the data will be collected from secondary sources, in this research from Thompson DataStream. Secondly, the data collected will be rearranged and calculated. The useful data will be selected for conducting tests. Thirdly, the selected data will be analyzed using E-Views 9. The last but not the least, we analyzed the outcome of the analysis.

3.2.6. Data Analysis

In this particular study, the researcher uses E-view 9 software as the empirical tool to compute the variables data into empirical results for e-amination. E-view 9 is an econometric software package which provides general statistical analysis, estimating and forecasting techniques which are essential in conducting this investigation. The methods that will be used to analyze the data in this research are as follow:

3.2.6.1. Unit Root Test

In stock markets, empirical research is based on time series data. A pre-requisite for designing meaningful results in time series analysis is to have a stationary data in order to enhance the accuracy and reliability of the models constructed. If the time series data is not stationary, regression parameters cannot be carried out, or if they are carried out the results may not be accurate. A time series data is considered as stationary if its mean and variance are constant over a given period of time, and covariance are constant for a given lag. The stronger is the stationary of the data, the best is for the research because it does not lead to spurious regression. One of the most common ways to test whether a time series data is stationary or not is using the unit root test. Although there are several unit root tests to check stationary data, this paper is using Augmented Dickey Fuller (ADF) test and Phillip-Perron (PP) test.

3.2.6.1.1. Augmented Dickey Fuller (ADF) Test

ADF test is the most basic unit root test in time series data developed by Dickey and Fuller in 1981. The reason to select this test is because it is efficient in e-amining the dynamic relationship between independent and depend variables. According to (Al-Zoubi & Al-Sharkas, 2011) ADF consists a running regression of the first difference of time series data against the time series data lagged once, lagged difference terms and optionally a constant and a time trend. Different literatures such as Asaolu & Ogunmuyiwa (2011), Karunanayake et al (2010), and Agrawal et al. (2010) used E-views to conduct ADF test in their studies to avoid spurious regression. Therefore, analyzing the stationary of the time series data is essential before conducting any test in the series. By applying regression model on non-stationary variables can give ambiguous parameter estimates of the relationships between the dependent and independent variables. The following ADF model is used to check whether the time series data is stationary or not:

Abbildung in dieser Leseprobe nicht enthalten

Where:

[Abbildung in dieser Leseprobe nicht enthalten]time series to be tested ܾ଴ = intercept term

[Abbildung in dieser Leseprobe nicht enthalten] coefficient of interest in the unit root test ο = difference operator

[Abbildung in dieser Leseprobe nicht enthalten] = parameter of the augmented lagged first difference of ܻ௧ ߳௧ = white noise error term

The unit root test is conducted on the coefficient of [Abbildung in dieser Leseprobe nicht enthalten]in the regression. If the coefficient is less than zero, we reject the hypothesis that y contains unit root. If we reject the null hypothesis, it means the series is stationary. The null and alternative hypothesis would be:

[Abbildung in dieser Leseprobe nicht enthalten]: Variable is not stationary or has unit root.

[Abbildung in dieser Leseprobe nicht enthalten]: Variable is stationary or has no unit root.

The decision rule for the ADF test is that we will reject [Abbildung in dieser Leseprobe nicht enthalten] if the probability value of the unit root test is less than the significance level. Otherwise, if the probability value of the unit root test is more than the significance level we do not reject [Abbildung in dieser Leseprobe nicht enthalten].

Phillips-Perron (PP) Test

The Phillips-Perron test (PP test) was introduced as an alternative to ADF test (Vejzagic & Zarafat, 2013). According to Onour (2009), PP test differs from ADF test mainly in the way how they deal with serial correlation and heteroskedasticity in errors. PP test uses a non-parametric statistical method and avoid the use of adding lagged difference terms.

The null and alternative hypothesis would be:

[Abbildung in dieser Leseprobe nicht enthalten]: Variable is not stationary or has unit root.

[Abbildung in dieser Leseprobe nicht enthalten] Variable is stationary or has no unit root.

The decision rule for the PP test is that we will reject [Abbildung in dieser Leseprobe nicht enthalten] if the probability value of the unit root test is less than the significance level. Otherwise, if the probability value of the unit root test is more than the significance level we do not reject [Abbildung in dieser Leseprobe nicht enthalten].

3.2.7. Diagnostic Checks

According to Sekar (2010) diagnostic checks have become a standard tool to identify models before forecasting the data. Time series models usually face econometric problems such as multicollinearity, heteroscedasticity and autocorrelation problem. Therefore, if all these problems can be avoided, the model is considered BLUE (Best, Linear, Unbiased and Estimator). The technique that can help to detect econometric problems in a time series data is called Ordinary Least Square (OLS). Therefore, this research will use OLS method to detect the economic problems affecting the time series data, and also ensure that the research is free from such economic problems.

3.2.7.1. Multicollinearity

Multicollinearity arises when there are multiple variables in a single regression model which are correlated and give redundant information. Thus, multicollinearity problem can mislead the test, making the test results of the model insignificant. According to Gujariti & Porter (2009), multicollinearity problem can arise from wrong usage of dummy variables for equation or including similar variable which are e-tremely correlated.

In order to detect the multicollinearity problem in the model, we will e-amine the correlation matri- and compute the Variance Inflation Factor (VIF). Pearson’s correlation analysis is used to detect any high pair wise correlation between independent variables. The Pearson’s correlation measurement scale varies from -1 to +1. Usually, the value of +1 represents a perfect positive relationship between the variables, which means the two variables are moving toward the same direction. On the other hand, the value of -1 represents a perfect negative relationship between the variables, which means the two variables are moving to opposite directions. According to Kuwornu (2012) in a correlation matri-, a range of -0.70 to +0.70 is acceptable. Thus, if correlation between the variables is higher than this range, multicollinearity problem may occur.

If a pair of independent variables is highly correlated to each other, an au-iliary regression model of the highly paired variables is developed to determine the severity of the multicollinearity. Thus, VIF is computed in order to detect such severity of the multicollinearity problem. The formula is given as:

Abbildung in dieser Leseprobe nicht enthalten

Where:

[Abbildung in dieser Leseprobe nicht enthalten] = Coefficient of determination

The null and alternative hypothesis would be:

[Abbildung in dieser Leseprobe nicht enthalten]

[Abbildung in dieser Leseprobe nicht enthalten]: VIF < 10 (multicollinearity problem does not e-ist) [Abbildung in dieser Leseprobe nicht enthalten] VIF > 10 (multicollinearity problem e-ist)

The decision rule for multicollinearity is that we do not reject [Abbildung in dieser Leseprobe nicht enthalten] if the VIF value is less than ten. On the other hand, if the value of VIF is more ten we reject [Abbildung in dieser Leseprobe nicht enthalten], meaning that there is a serious multicollinearity problem.

3.2.7.2. Heteroskedasticity

A time series regression should consist of same variance of distribution. Therefore, if the variance of distribution is not the same, it violates the assumptions that the variances of the error terms are constant, giving rise to heteroskedasticity problem. Heteroskedasticity can be caused by different factors such as missing an e-planatory variable or the variables are not normally distributed. White (1980) argued that heteroskedasticity affects the efficiency of estimated parameter and covariance matri-, misleading the results of the hypothesis testing. Furthermore, Long & Laurie (1998) argued that heteroskedasticity problem in time series data tend to underestimate the variances and standard errors, leading results of both t statistics and F-statistic to be unreliable. Engle (1982) established the Autoregressive Conditional Heteroskedasticity (ARCH) test to detect heteroskedasticity problem in time series data. The hypotheses for heteroskedasticity test are as follow:

[Abbildung in dieser Leseprobe nicht enthalten]: There is no heteroskedasticity problem.

[Abbildung in dieser Leseprobe nicht enthalten] There is heteroskedasticity problem.

The decision rule for heteroskedasticity test is that we reject [Abbildung in dieser Leseprobe nicht enthalten] if the probability value is less than significance level. Otherwise, if the probability value is more than the significance level we do not reject [Abbildung in dieser Leseprobe nicht enthalten]Ǥ

3.2.7.3. Autocorrelation

Autocorrelation can be referred as the measurement of correlation coefficient. Therefore, if there is autocorrelation in the series, it means that the error terms of the observations are related to each other, which violates the assumptions of Classical Linear Regression Model (CLRM) that the regressions must fulfill. Autocorrelation analysis is employed to determine the performance of data in a time series model. In order to determine any correlation between the independent variables, Pearson’s correlation analysis can also be employed.

Correlation is a statistical method used to measure the relationship between variables. The scale varies from -1 to +1. -1 represents negative correlation between the variables, +1 represents positive correlation between the variables and 0 represents no correlation at all. It is usually normal for time series data to face autocorrelation problems. According to Gujarati & Porter (2009) there are two methods used to detect autocorrelation problem namely; Durbin-Watson and Breusch-Godfrey LM.

The Durbin-Watson statistics ranges from 0 to 4. A value near 2 indicates non-autocorrelation; a value toward 0 indicates positive autocorrelation; a value toward 4 indicates negative autocorrelation. The Durbin-Watson statistics model is represented as follow:

Abbildung in dieser Leseprobe nicht enthalten

Where:

[Abbildung in dieser Leseprobe nicht enthalten] = residual for the time period t

[Abbildung in dieser Leseprobe nicht enthalten]= residual for the preceding time period (t-1)

However, in this study in order to detect the autocorrelation problem, the Breusch-Godfrey LM test will be conducted. The following are the null and alternative hypothesis for the BreuschGodfrey LM test:

[Abbildung in dieser Leseprobe nicht enthalten]: There is no autocorrelation problem.

[Abbildung in dieser Leseprobe nicht enthalten] : There is an autocorrelation problem.

The decision rule for Breusch-Godfrey LM test is that we rejects [Abbildung in dieser Leseprobe nicht enthalten] if probability value is less than significance level. Otherwise, if probability value is more than significance level, we do not reject [Abbildung in dieser Leseprobe nicht enthalten].

3.2.7.4. Model Specification

Model specification is correct when the relevant independent variables are chosen and included in the model, and when appropriate functional form of variable into the model is selected (Gujarati & Porter, 2009).

The omission of relevant variables occurs when the model does not include e-planatory variables that are important in e-plaining the dependent variable. On the other hand, the inclusion of irrelevant variables occurs when the model includes variables that do not have a significant impact in e-plaining the dependent variable. In addition, the model might be incorrectly specified when it includes wrong functional form of variables. It is typically e-perienced when an assumption of a linear equation is made where the relationship is non-linear.

Therefore, when irrelevant independent variables are chosen and selected, they are correlated with error term, which will provide biased results. Hence, if the model includes irrelevant variables and omits the relevant variables, and also includes wrong functional form of variables, the model becomes incorrectly specified, providing misleading results. In this research, the model specification problem can be detected by conducting the Ramsey RESET test, developed by Ramsey in 1969. The null and alternative hypotheses for this test are as follow:

[Abbildung in dieser Leseprobe nicht enthalten]: Model specification is correct.

[Abbildung in dieser Leseprobe nicht enthalten] Model specification is incorrect.

The decision rule for model specification is that we reject [Abbildung in dieser Leseprobe nicht enthalten] when probability value is less than the significance level. Otherwise, if the probability value is more than the significance level we do not reject [Abbildung in dieser Leseprobe nicht enthalten].

3.2.7.5. Normality Test

Normality test is conducted to investigate the normality of the residual term. Researchers can use histogram of the residual to test the normality’s assumption of error terms. Usually, if the residual term is normally distributed, the specification model is also correct. A normality test can be conducted using the Jarque-Bera (JB) test statistic developed by Carlos Jarque and Anil K. Jarque-Bera in 1987 (Jarque & Bera, 1987). The test statistics of Jarque-Bera can be computed using the following formula:

Abbildung in dieser Leseprobe nicht enthalten

Where:

N = sample size

S = Skewness coefficient

K = Kurtosis coefficient

The hypotheses for this test are as follow:

[Abbildung in dieser Leseprobe nicht enthalten]: The error term is normally distributed.

[Abbildung in dieser Leseprobe nicht enthalten] The error term is not normally distributed.

Abbildung in dieser Leseprobe nicht enthalten

The decision rule for the Jarque-Bera test statistic is that we reject [Abbildung in dieser Leseprobe nicht enthalten] if the probability value is less than significance level. On the other hand, if the probability value of [Abbildung in dieser Leseprobe nicht enthalten] is more than the significance level, we do not reject [Abbildung in dieser Leseprobe nicht enthalten].

3.2.8. Granger Causality Test

Granger (1969) developed granger causality test in order to determine causal relationship between two variables and e-amine whether one time series data is significant in forecasting another (Harasheh & Libdeh, 2011). Granger causality tests aims to e-amine whether the past values of a variable can be significant in forecasting changes in another variable. Granger (1969) argued that granger causality is a suitable test to determine the interaction between movements of stock price and economic changes.

The granger causality test is used to determine the short run relationship between the dependent and independent variables. This test provides two outcomes which are unidirectional and bidirectional causality between the variables. Kutty (2010) used granger causality test to determine the e-istent relationship between e-change rates and Me-ico’s stock prices. However, the current research aims to investigate whether macroeconomic variables are significant factors in e-plain Malaysian stock market returns. Additionally, this test can only be conducted when the data is in stationary form. Therefore, Granger causality test is performed after the ADF or PP tests have been performed. The test is based on the following equations:

Abbildung in dieser Leseprobe nicht enthalten

Where:

[Abbildung in dieser Leseprobe nicht enthalten]variables to be tested

[Abbildung in dieser Leseprobe nicht enthalten] = uncorrelated errors

[Abbildung in dieser Leseprobe nicht enthalten] = time period

[Abbildung in dieser Leseprobe nicht enthalten] & ݈ = number of lags

The null and alternative hypothesis would be: [Abbildung in dieser Leseprobe nicht enthalten]: X does not granger cause Y.

[Abbildung in dieser Leseprobe nicht enthalten] X does granger cause Y.

The decision rule for granger causality test is that we reject [Abbildung in dieser Leseprobe nicht enthalten] if the probability value is less than the significance level. Otherwise, if the probability value is more than the significance level, we do not reject [Abbildung in dieser Leseprobe nicht enthalten].

3.2.9. Multiple Regressions Model

Multiple regressions model is a method of analyzing data to determine the linkage between dependent and independent variables. In addition, this model can also identify econometric problems such as heteroscedasticity, autocorrelation and multicollinearity. In order to determine the relationship the dependent and independent variables OLS regression model will be applied. The dependent variable is the stock market, and the independent variables are e-change rates, inflation (CPI), crude oil prices and foreign direct investment. The functional method of this model is described below:

Abbildung in dieser Leseprobe nicht enthalten

Where:

KLCI = Stock market returns

β0 = Intercept term

ER = E-change rate

CPI = Pro-y of inflation rate

COP = Crude oil price

FDI = Foreign Direct Investment

e = Residual error

The regression model is essential to determine the significance of each independent variable towards the dependent variable. The β0 is a constant; ER, CPI, COP, and FDI are considered as coefficients and e the error term.

3.2.9.1. T-test

The t-test statistic was developed by William Sealy Gosset (1908). This test is conducted in order to determine whether independent variables individually have a significant impact on the dependent variable or not. In this study, the independent variables are e-change rate, inflation, crude oil price and FDI, and the dependent variable is the KLCI. The t-test is based in the assumption that the samples drawn are independently, the sample size is very small, usually not more than 30 and that error terms are normally distributed. This test can be computed using the following formula:

Abbildung in dieser Leseprobe nicht enthalten

The hypotheses for this test are the following:

[Abbildung in dieser Leseprobe nicht enthalten]: There is no significant relationship between the independent and dependent variable.

[Abbildung in dieser Leseprobe nicht enthalten] There is a significant relationship between the independent and dependent variable.

The decision rule for the t-test is that we reject [Abbildung in dieser Leseprobe nicht enthalten] if the probability value is less than the significance level. Otherwise, if the probability value is more than the significance level, we do not reject [Abbildung in dieser Leseprobe nicht enthalten]Ǥ

3.2.9.2. F-test

F-test was developed by Ronald Aylmer Fisher (1924), and is used to measure the overall significance of the estimated multiple regression models. The features and assumptions of this test are similar with those of the t-test. The statistic value of the F-test can be computed by applying the following formula:

The null and the alternative hypotheses for this test are as follow:

[Abbildung in dieser Leseprobe nicht enthalten]: The overall model is insignificant.

[Abbildung in dieser Leseprobe nicht enthalten] The overall model is significant.

The decision rule for F-test is that we reject [Abbildung in dieser Leseprobe nicht enthalten] if the probability value is less than significance level. On the other hand, if the probability value is more than significance level, we do not reject [Abbildung in dieser Leseprobe nicht enthalten]Ǥ

3.2.10. Research Hypothesis

The research aims to investigate whether the variables are correlated or not.

Ho: the variables are not correlated.

H1: the variables are correlated.

Table 2: Research Hypothesis

Abbildung in dieser Leseprobe nicht enthalten

3.3. Conclusion

This chapter discusses in deep the methodologies and statistical analysis that will be conducted in order to test the research hypothesis and determine whether there is a relationship between stock market returns in Malaysia and macroeconomic variables or not.

Chapter 4: Data Analysis

4.0. Introduction

This chapter analyses and interprets the empirical results from the methodologies described in the previous chapter which are applied in this research. The data used in this paper was collected from Thomson DataStream, and analyzed using the E-views 9. The tests conducted in this chapter include Unit Root Test, multicollinearity, heteroskedasticity, autocorrelation, model specification, normality test, Granger Causality test and multi-regression model. The significance level (α) for all the tests is 0.05. The results obtained will be e-pressed in table form and a clear e-planation will be provided based on the findings.

4.1. Unit Root Test

The Unit Root Test is conducted in order to identify the stationary of the data. In this case, if the data is not stationary, regression parameters cannot be carried out. Therefore, to check the stationary of the data, Augmented Dickey Fuller (ADF) test and Phillips-Perron (PP) test will be conducted.

Hypothesis

[Abbildung in dieser Leseprobe nicht enthalten] : lnKLCI / lnER / lnCPI / lnCOP / lnFDI is not stationary and has a unit root.

[Abbildung in dieser Leseprobe nicht enthalten]: lnKLCI / lnER / lnCPI / lnCOP / lnFDI is stationary and does not have a unit root.

α = 0.05

Decision rule

1. Reject [Abbildung in dieser Leseprobe nicht enthalten] if probability value is less than significant level (α = 0.05), meaning that the variable is stationary.

2. Do no reject [Abbildung in dieser Leseprobe nicht enthalten] if probability value is more than significant level (α = 0.05), meaning that the variable is not stationary.

4.1.1. Augmented Dickey-Fuller (ADF) Test

Table 3: Results of ADF Test

Abbildung in dieser Leseprobe nicht enthalten

Source: Developed from E-views 9.

The level stage:

Constant: at the level stage the [Abbildung in dieser Leseprobe nicht enthalten] for all the variables at 5% significance level is not rejected. Therefore, all the variables are not stationary at level stage.

Constant with trend: at the level stage the [Abbildung in dieser Leseprobe nicht enthalten] for all the variables e-cept for lnCPI at 5% significance level is not rejected. Therefore, all the variables e-cept for lnCPI are not stationary at level stage.

The first difference:

Constant: at first difference the [Abbildung in dieser Leseprobe nicht enthalten] for all the variables at 5% significance level is rejected. Therefore, all the variables are stationary at first difference.

Constant with trend: at first difference the [Abbildung in dieser Leseprobe nicht enthalten] for all the variables at 5% significance level is rejected. Therefore, all the variables are stationary at first difference.

4.1.2. Phillips Perron (PP) Test

Table 4: Results of PP Test

Abbildung in dieser Leseprobe nicht enthalten

Source: Developed from E-views 9.

The level stage:

Constant: at the level stage the [Abbildung in dieser Leseprobe nicht enthalten] for all the variables at 5% significance level is not rejected. Therefore, all the variables are not stationary at level stage.

Constant with trend: at the level stage the [Abbildung in dieser Leseprobe nicht enthalten] for all the variables at 5% significance level is not rejected. Therefore, all the variables are not stationary at level stage.

The first difference:

Constant: at first difference, the [Abbildung in dieser Leseprobe nicht enthalten] for all the variables at 5% significance level is rejected. Therefore, all the variables are stationary at first difference.

Constant with trend: at first difference, the [Abbildung in dieser Leseprobe nicht enthalten] for all the variables at 5% significance level is rejected. Therefore, all the variables are stationary at first difference.

4.2. Diagnostic Checks

4.2.1. Multicollinearity

Multicollinearity is the problem that occurs when there is a relationship among independent variables, which can mislead the test results.

Hypothesis

[Abbildung in dieser Leseprobe nicht enthalten] : There is no multicollinearity problem.

[Abbildung in dieser Leseprobe nicht enthalten] : There is a multicollinearity problem.

Decision rules

1. We do not reject [Abbildung in dieser Leseprobe nicht enthalten] if VIF < 10, meaning that there is no multicollinearity problem.

2. We reject [Abbildung in dieser Leseprobe nicht enthalten] if VIF > 10, meaning that there is a serious multicollinearity problem.

Table 5: Correlation Test of Variables

Abbildung in dieser Leseprobe nicht enthalten

Source: Developed via E-views 9.

Based on Person’s correlation analysis, the data in the matri- shows that there is one pair of independent variables which is highly correlated to each other, which is lnCPI and lnFDI of - 0.734780. Therefore, we will conduct a regression analysis of the correlated pair (lnCPI and lnFDI) to get the R-Square ([Abbildung in dieser Leseprobe nicht enthalten]) and determine the Variance Inflation Factor (VIF). VIF is computed in order to determine the presence of the multicollinearity problem in the model.

Figure 3: Regression Result of lnCPI and lnFDI

Abbildung in dieser Leseprobe nicht enthalten

Source: Developed via E-views 9.

Abbildung in dieser Leseprobe nicht enthalten

From the VIR result, it can be concluded that the model does not have multicollinearity problem between the variables lnCPI and lnFDI since the VIF = 1.1519, which is less than 10.

4.2.2. Heteroskedasticity

The heteroskedasticity problem refers to the inconsistent characteristic of variances in the time series data. Therefore, if the variance of distribution is not consistent, heteroskedasticity problem may e-ist. In this research, Autoregressive Conditional Heteroscedasticity (ARCH) test is carried out to investigate the problem of heteroscedasticity in the series.

Hypothesis

[Abbildung in dieser Leseprobe nicht enthalten] : There is no heteroskedasticity problem in the model. [Abbildung in dieser Leseprobe nicht enthalten] : There is a heteroskedasticity problem in the model.

α = 0.05

Decision rules

1. We reject [Abbildung in dieser Leseprobe nicht enthalten] if the P-value of the Chi-squared is less than significance level, meaning that there is heteroskedasticity problem.

2. We do not reject [Abbildung in dieser Leseprobe nicht enthalten] if the P-value of the Chi-squared is more than significance level, meaning that there is no heteroskedasticity problem.

Table 6: Heteroskedasticity Test (ARCH)

Abbildung in dieser Leseprobe nicht enthalten

Source: Developed via E-views 9.

Conclusion

Since the P-value of the Chi-squared obtained (0.2564) is greater than the significance level (α = 0.05), we do not reject [Abbildung in dieser Leseprobe nicht enthalten]Ǥ Therefore, we have enough reason to conclude that there is no heteroskedasticity problem in the model.

4.2.3. Autocorrelation

This research paper uses the Breusch-Godfrey LM test to detect the presence of autocorrelation problem in the model.

Hypothesis

[Abbildung in dieser Leseprobe nicht enthalten]Ǥ: There is no autocorrelation problem in the model. [Abbildung in dieser Leseprobe nicht enthalten]Ǥ: There is autocorrelation problem in the model.

α = 0.05

Decision rules

1. We do not reject [Abbildung in dieser Leseprobe nicht enthalten]Ǥ if P-value of the Chi-squared is more than significance level, meaning that there is no autocorrelation problem.

2. We reject [Abbildung in dieser Leseprobe nicht enthalten]Ǥ if P-value of the Chi-squared is less than significance level, meaning that there is an autocorrelation problem.

Table 7: Breush-Godfrey Serial Correlation LM Test

Abbildung in dieser Leseprobe nicht enthalten

Source: Developed via E-views 9.

Conclusion

Since the P-value of Chi-Squared obtained (0.5804) is more than the significance level (α = 0.05), we do not reject [Abbildung in dieser Leseprobe nicht enthalten]Ǥ Thus, we can conclude that there is no autocorrelation problem in the model.

4.2.4. Model Specification

In this research, model specification problem can be detected by conducting the Ramsey RESET test.

Hypothesis

[Abbildung in dieser Leseprobe nicht enthalten]Ǥ: The model is correctly specified.

[Abbildung in dieser Leseprobe nicht enthalten]Ǥ: The model is not correctly specified.

α = 0.05

Decision rules

1. Do not reject [Abbildung in dieser Leseprobe nicht enthalten] if P-value of F-statistic is more than the significance level, meaning that the model is correctly specified.

2. Reject [Abbildung in dieser Leseprobe nicht enthalten] if P-value of F-statistic is less than the significance level, meaning that the model is not correctly specified.

Table 8: Ramsey RESET test Output

Abbildung in dieser Leseprobe nicht enthalten

Source: Developed via E-views 9.

Conclusion

Since the P-value of F-statistic obtained (0.4263) is greater than the significance level (α = 0.05), we do not reject [Abbildung in dieser Leseprobe nicht enthalten]Ǥ Therefore, we can conclude that the model is correctly specified.

4.2.5. Normality Test

In this research paper, the normality error term is determined using Jarque-Bera test.

Hypothesis

[Abbildung in dieser Leseprobe nicht enthalten]: Error term is normally distributed.

[Abbildung in dieser Leseprobe nicht enthalten] Error term is not normally distributed.

α = 0.05

Decision rules

1. Do not reject [Abbildung in dieser Leseprobe nicht enthalten] if the P-value of Jarque-Bera test statistic is more than the significance level, meaning that the error term is normally distributed.

2. Reject [Abbildung in dieser Leseprobe nicht enthalten] if the P-value of Jarque-Bera test level is less than the significance level, meaning that the error term is not normally distributed.

Figure 4: Jarque-Bera Normality Test

Abbildung in dieser Leseprobe nicht enthalten

Source: Developed via E-views 9.

Conclusion

Since the P-value of Jarque-Bera (000000) is less than the significance level (α = 0.05), we reject [Abbildung in dieser Leseprobe nicht enthalten]Ǥ Therefore, the error term is not normally distributed.

4.3. Granger Causality Test

Granger Causality Test is conducted in this study in order to determine the causal relationship between the variables and the KLCI. According to Ali et al. (2010) if the causal relationship between variables e-ists, they can be used to forecasting changes of each other.

Hypothesis

[Abbildung in dieser Leseprobe nicht enthalten] : X does not granger cause Y. [Abbildung in dieser Leseprobe nicht enthalten] : X does granger cause Y.

α = 0.05

Decision rule

1. Reject [Abbildung in dieser Leseprobe nicht enthalten], if probability value is less than the significant level.

2. Do not reject [Abbildung in dieser Leseprobe nicht enthalten] , if probability value is more than the significant level.

Table 9: Granger Causality Test Results

Abbildung in dieser Leseprobe nicht enthalten

Source: Developed via E-views 9.

Conclusion

The results showed in the table 4.9 suggest that all the independent variables, e-change rate, inflation, crude oil price and FDI do not granger cause KLCI at 5% confidence level.

4.4. Regression Model

After performing the diagnostic tests of the model, the ne-t step is to determine the empirical model for the series. OLS regression model is used in this research to determine the relationship between stock market returns and the selected macroeconomic variables in Malaysia.

[Abbildung in dieser Leseprobe nicht enthalten]

Where:

[Abbildung in dieser Leseprobe nicht enthalten] = the first difference of the natural logarithm of Kuala Lumpur Composite Inde- (KLCI) stock price.

[Abbildung in dieser Leseprobe nicht enthalten] = the first difference of the natural logarithm of Foreign E-change Rate (FX). ݈݀݊ሺܥܲܫ) = the first difference of the natural logarithm of Consumer Price Inde- (CPI). ݈݀݊ሺܥܱܲሻ = the first difference of the natural logarithm of Brent Crude Oil Price (COP).

݈݀݊ሺܨܦܫሻ = the first difference of the natural logarithm of inflow of Foreign Direct Investment

(FDI) to Malaysia.

Table 10: OLS Regression Result of the Model

Abbildung in dieser Leseprobe nicht enthalten

Source: Developed via E-views 9.

dln(KLCI) = 0.008803 - 0.622279 dln(ER) - 2.020344dln(CPI) + 0.132935dln(COP) + 0.001386dln(FDI) + e

The determinant of coefficient (R-squared) is used to measure the percentage of variation in dependent variable e-plained by total variation of independent variables. Based on the table 4.7, R-squared (0.305947) indicates that 30.59% of variation in Malaysian stock market returns is e-plained by the total variation in e-change rate, inflation rate, crude oil prices and FDI.

4.4.1. T-test

Hypothesis

[Abbildung in dieser Leseprobe nicht enthalten]: βi = 0 (insignificant)

[Abbildung in dieser Leseprobe nicht enthalten] βi ≠ 0 (significant) where i = ER, CPI, COP, FDI

α = 0.05

Decision rules

1. Reject [Abbildung in dieser Leseprobe nicht enthalten] if probability value of test statistic is less than significance level, meaning that there is a significant relationship between the independent and dependent variables.

2. Do not reject [Abbildung in dieser Leseprobe nicht enthalten] if probability value of test statistic is more than significance level, meaning that there is no a significant relationship between the independent and dependent variables.

Table 11: T- test Results

Abbildung in dieser Leseprobe nicht enthalten

Source: Developed via E-views 9.

Conclusion

The t- test of OLS regression model indicates that e-change rate, inflation rate and crude oil price have a significant impact on the Malaysian stock market returns. However, FDI does not have any influence on the stock market returns in Malaysia.

4.4.2. F-test

Hypothesis

[Abbildung in dieser Leseprobe nicht enthalten]: The overall model is insignificant. [Abbildung in dieser Leseprobe nicht enthalten] The overall model is significant.

α = 0.05

Decision rules

1. Reject [Abbildung in dieser Leseprobe nicht enthalten] if probability value of F-statistic value is less than significance level, meaning that the overall model is significant.

2. Do not reject [Abbildung in dieser Leseprobe nicht enthalten] if probability value of F-statistic value is more than significance level, meaning that the overall model is insignificant.

Table 12: F- test Results

Abbildung in dieser Leseprobe nicht enthalten

Source: Developed via E-views 9.

Conclusion

Since the F-statistic value (0.000000) is less than the significance level (α = 0.05), we reject [Abbildung in dieser Leseprobe nicht enthalten]. Therefore, we can conclude that the model is significant in e-plaining the stock market returns in Malaysia.

4.5. Conclusion

This chapter has carried out different tests which are crucial when analyzing time series data. Unit root test was carried out in order to ensure the stationarity of the data; multicollinearity, heteroscedasticity, autocorrelation, model specification and normality tests were conducted to verify whether the model is free from econometric problems whose e-istence reduces the reliability of the results; Granger causality test was conducted to determine the short run relationship between the dependent and independent variables; and finally the OLS regression model was also carried out to determine the long run dynamic relationship between the variables. The results of all these tests are represented in figure and table form. These results, the major findings and the implications will be discussed further in the ne-t chapter.

Chapter 5: Findings, Discussions and Implications

5.0. Introduction

The chapter five of this research paper presents the summary of the overall results of the impact of selected macroeconomic variables on stock market returns in Malaysia. First and foremost, this chapter presents the overall statistical analyses, which is basically the summary of the overall results of the previous chapter. Ne-t, major findings discuss about the dynamic relationship between the dependent and independent variables. The third part provides the implications of that this study may have for different parties who are interested in understanding the movement of the stock market in Malaysia.

5.1. Summary of Unit Root Test

Statistically it has been established that many of macroeconomic variables such as stock inde-, inflation rate, interest rate, gross domestic product (GDP) and other variables tend to behave in a non-stationary manner. This means that these variables tend to increase and decrease over the time. However, when analyzing time series data, it is essential that this type of data is in stationary form. Therefore, in conjunction with other studies such as Agrawal et al. (2010) and Sari & Soytas (2006), this study conducted unit root tests (ADF and PP) to test for the stationarity of the data. Testing the stationarity of the data is important to ensure that the statistical properties of the variables are constant over time.

In this study, the unit root tests ADF and PP were conducted in order to ensure the stationarity of the data. At 5% confidence level, both tests ADF & PP found that all the variables (KLCI, ER, CPI, COP and FDI) have unit roots, meaning that they were non-stationary at level. This is because the absolute value of the test statistics for every variable was lesser than the absolute variables of the test critical values at 5% confidence level. Thus, in order to ensure stationarity of the data, all the variables had to be transformed into first difference. After performing the first difference for all the variables, results showed that for both ADF & PP tests, at 5% confidence level all the variables did not have unit roots, meaning that they were stationary. Therefore, we can conclude that at 5% confidence level, KLCI, ER, CPI, COP and FDI are stationary.

5.2. Summary of Diagnostic Tests

Table 13: Summary of Diagnostic Tests

Abbildung in dieser Leseprobe nicht enthalten

This study performed several diagnostic tests to ensure that the data used was consistent with all the OLS assumptions. These tests include multicollinearity, heteroskedasticity, autocorrelation, model specification and normality tests. The rationale behind performing all these tests is because time series data are non-stationary by nature and as a result they tend to provide biased results. Hence, there is a need to perform a number of diagnostic tests to ensure its stationarity and reliability when testing the hypotheses.

5.2.1. Multicollinearity

The results of multicollinearity test in this study did not show any high correlation between the independent variables, e-cept for inflation and FDI (-0.734780). Therefore, in order to determine the e-istence of multicollinearity problem in the model, the study computed the VIF in the highly correlated pair (ER and FDI). The VIF of the highly correlated pair showed 1.1519, which is very low compared to the VIF value of 10. If the VIF value was greater than 10, it would mean the e-istence of multicollinearity problem in the model. However, VIF value is less than 10, which indicates that the model is free from multicollinearity problem.

5.2.2. Heteroskedasticity

According to Hail & Asteriou (2014) heteroskedasticity problem causes the regression model to underestimate the variances and standard errors, leading to higher than e-pected F-statistics and t-statistics values. Therefore, if there is heteroskedasticity problem in the model, F-statistics and t-statistics are no longer reliable to test the hypothesis because they lead us to reject the null hypothesis too often. The Chi-Square probability value of the Breusch-Pagan-Godfrey Heteroskedasticity test shows 0.2564, which is more than the significance level (0.05). This means that the null hypothesis that residuals are not heteroskedastic cannot be rejected. Therefore, the residuals are homoskedastic. It is important to note that heteroskedasticity is performed using stationary data. Therefore, in this study we use the first difference time series data to conduct this test, since the data is non-stationary at level.

5.2.3. Autocorrelation

Using first difference time series data at 5% confidence level, the results showed no serial correlation among the regression residuals. The Chi-Square probability value of the Breusch- Godfrey Serial Correlation LM Test obtained was 0.5604, which is greater than the significance level (0.05). This leads to the conclusion that the null hypothesis of no serial correlation among the regression variables cannot be rejected. Therefore, the regression residuals do not contain autocorrelation problem. Consequently, the OLS regression coefficients are unbiased.

5.2.4. Model Specification

In conjunction with the study of Din & Mubasher (2013), this study used OLS method to compute the estimates of the regression model. Din & Mubasher (2013) used the same model to e-amine the effect of macroeconomic variables on the stock price movement in Indian stock market. The P-value of F-statistic (0.4263) obtained from Ramsey RESET test which is greater than the significance level (0.05), indicates that the model is correctly specified. Therefore, we can conclude that all the variables are important in e-plaining the variation of the stock market returns in Malaysia, and the regression model is the right functional form of the variables.

5.2.5. Normality

The study also carried out the normality test to determine whether the residuals are normally distributed or not using Jarque-Bera test statistic. The P-value of Jarque-Bera test statistic obtained (0.00000) is lesser than the significance level (0.05). Therefore, the residuals are not normally distributed. The abnormal distribution of residuals in the OLS regression model is an obvious violation of the CLRM assumptions, which means that t-statistics and F-statistics would provide biased results. However, according to Wooldridge (2015) even if the CLRM assumptions do not hold, OLS is still consistent, and the usual inference procedures are valid as long as the other assumptions are valid. Thus, since the other CLRM assumptions hold, OLS is still a consistent method to determine the impact of macroeconomic variables on stock market returns in Malaysia.

5.3. Summary of Granger Causality Test

Granger causality test was also conducted to establish the unidirectional and/or bidirectional relationship between the dependent and independent variables. On the basis of Granger causality test, e-change rate, inflation, crude oil price and FDI do not granger cause KLCI. This means that in the short run, these macroeconomic variables do not affect the performance of the Malaysian stock market. However, results show that there is a unidirectional causal relationship from KLCI to crude oil prices. This result suggests that past values of KLCI could be used to predict future values of crude oil. Therefore, on the basis of this evidence, if KLCI values increased in the past, we can predict that crude oil prices will rise in the future.

5.4. Summary of OLS Regression

Figure 5: OLS Regression Results

Abbildung in dieser Leseprobe nicht enthalten

Source: Developed via E-views

The figure above illustrates the relationship between Malaysia stock market returns and selected macroeconomic variables. Based on the above results, the R-squared (0.305947) is relatively low, which implies a low forecasting power of the multivariate model, and thus, there is a high likelihood of the e-clusion of relevant macroeconomic variables which might significantly affect the variation of stock market returns in Malaysia. The F-statistic value (0.000000) is also very low, which implies that the selected macroeconomic variables jointly do not have a significant impact on Malaysia stock market performance. Moreover, results show that all the variables crude oil prices, e-change rate and inflation e-cept FDI are individually significant in e-plaining the variation of stock market returns. Crude oil prices and FDI are found to have a positive relationship with Malaysia stock market, whereas inflation and e-change rate have a negative impact on the stock market in Malaysia.

5.5. Discussion on Major Findings

This study investigates the dynamic relationship between the Malaysian stock market returns and four selected macroeconomic variables. The R-squared value from the OLS regression indicates that e-change rate, inflation, crude oil prices and FDI are accountable for only 30.59% of the total variation of Malaysian stock market returns, which means that 69.41% variation in the stock market returns in Malaysia, are e-plained by other variables. Therefore, this result suggests that some important macroeconomic variables such as interest rate, money supply, GDP and other variables which might have a huge impact on the stock market returns have been e-cluded from this study. Moreover, the probability value of the F-statistic is very small, which indicates that all the variables combined together do not have a significant effect on the variation of Malaysia stock market returns. According to Din & Mubasher (2013) this might also be due to the e-clusion of relevant predictive variables of the total variation of the stock market returns.

OLS regression results suggest that in the long run, there is a significant negative relationship between the e-change rate and KLCI. This result is in line with previous research conducted by Bello (2013), Fang & Miller (2002), Khan et al. (2012), Ouma & Muriu (2014) and Jawaid & Haq (2012). Khan et al. (2012) studied the impact of e-change rate on Karachi stock e-change. Findings suggest that an increase in e-change rate causes a decrease in the stock market returns because when foreign investors invest their money in the stock market, an increase in the e-change rates will decrease their income since they receive lesser amount in their own currency due to the depreciation of the local currency. Besides that, Fang & Miller (2002) have argued that the depreciation on the domestic currency increases the returns on dollar assets. Therefore, investors tend to shift their funds from the domestic assets such as stocks to dollar based assets for higher e-pected returns. This shift in portfolio composition favors dollar assets over domestic stocks, leading to a decline in the stock market prices and thus its returns as well. However, findings from this study are not supported by other studies like Cakan & Ejara (2013), Soenen & Hennigar (1998), Chiang, Yang & Wang (2000) and Yau & Nieh, (2009) who found a positive relationship between the e-change rate and stock market returns. Cakan & Ejara (2013) have pointed out that depreciation of the local currency makes e-porting cheaper which increases demand and sales for local products which may increase companies’ profitability. However, when the currency appreciates, e-ports decrease together with the profits of companies, consequently the stock price.

The second variable to be discussed is the inflation which is pro-ied by CPI. In the long run, inflation is found to have a strong negative impact on the stock market returns in Malaysia. This result is further supported by various previous literatures that found a negative relationship between the stock market returns and inflation such as Phuyal (2016), DeFina (1991), Humpe & Macmillan (2007), Aliyu (2011) and Qamri, Haq & Akram (2015). DeFina (1991) has argued that an increase in the inflation negatively affect the corporate income since it causes an immediate rise of cost, and slowly reducing output and consequently the share price. Moreover, Aliyu (2011) has stated that when there is prices uncertainty in the domestic economy, the volatility of nominal asset returns should reflect the volatility of CPI. Therefore, maintain a stable level of inflation in the country will hugely contribute to a growing stock market by increasing the number of foreign as well as local capital inflows. In fact, this research did not come across literatures that have pointed out a positive impact of inflation on the stock market returns either in Malaysia or in other countries.

Furthermore, crude oil prices have become a strong indicator in determining the variations in the stock market. OLS regression suggests a strong positive relationship between crude oil prices and stock market returns in Malaysia. This result is consistent with prior studies conducted by Mohanty et al. (2011), Jafarian & Safari (2015), Mollick & Assefa (2013). However, studies on the impact of crude oil prices on stock market are too diverse and deeper research still need to be carried out to determine the relationship between these two variables. This happens because when oil prices increase, Malaysian oil and gas companies’ profit also increase. However, it can negatively affect the non-oil and gas companies by increasing the cost of production, consequently increasing input cost, resulting in higher prices and thus, a drop in demand. Le & Chang (2011) carried out an investigation on the impact of crude oil prices on stock market in emerging and developed countries including Malaysia. They found that Malaysia’s stock market responds negatively to changes in crude oil price. Other researches which found an inverse relationship between Malaysia stock market and crude oil prices include Najaf & Najaf (2016), and Najaf (2016). Moreover, Maghyereh (2004) found very weak evidence on the response of emerging stock markets to oil price shocks including Malaysia. Therefore, there is still a need for advanced studies on the impact of oil prices on stock market, especially for countries like Malaysia which are hugely dependent on the oil e-ports.

Last but not least variable to be discussed is foreign direct investment. Based on OLS regression result, there is a positive and insignificant relationship between FDI and stock market returns in the long run. Therefore, FDI cannot be used as a predictor in the variation of stock market returns in Malaysia. In line with this result are the studies conducted by Musa & Ibrahim (2014) and Olugbenga & Grace (2015). In addition, Onyinyechi & Ekwe (2016) and Idenyi et al. (2016) investigated the impact of FDI in Nigerian stock market. In both studies, results show that FDI insignificantly affects the performance of the stock market in Nigeria, though the relationship between the two variables was found to be negative. Ali (2014) and Arcabic et al. (2013) also have analyzed the impact of FDI on stock market returns in Pakistan and Croatia respectively. Although both studies have also found a positive relationship between FDI and stock market returns, they have pointed this relationship to be highly significant. Arcabic et al. (2013) have argued that FDI stimulates rapid technological development and economic growth through the transfer of know-how and technology, and then in the long run FDI affects the stock market growth as well.

5.6. Implication of the Study

The current paper provides analyses on the performance of Malaysian stock market based on monthly data from January 2007 to December 2015. Findings from this study can provide a wide range of parties such as investors, government, policymakers, academicians and researchers with good insights about the equity market in Malaysia. From this research, people will have a better understanding on the relationship between macroeconomic variables and the stock market, and learn how the movement of these factors affects the performance of the equity market. Thus, this paper can be used as a guideline by every party who is interested in the equity market in Malaysia.

For the investors, findings from this paper can be used as an important tool to construct a favorable portfolio investment through an in-depth analysis on the trends of the selected stock market returns predictors namely e-change rate, inflation rate, crude oil prices and foreign direct investment. With more availability of information on the relevant factors that affect the stock market, the investors can easily weight the risk and return more accurately. Therefore, for either Malaysian or foreign investors who are interested in investing in the Malaysian stock market but they are simultaneously concerned about the volatility of stock market movement, the current paper can be used as a guideline to overcome such concerns.

For instance, investors should be alert of the inverse relationship between inflation and stock market. This happens because inflation reduces the purchasing power of customers. With a drop in the purchase power, lesser money would be available for investment. Hence, investors would sell their stocks which lead to a drop in the stock market prices due to higher supply than demand. Therefore, appropriate policies should be implemented to contain inflation for a healthy stock market which plays an important role in the development of the economy.

Knowing the inverse relationship that e-change rate and inflation have with the stock market, the policymakers and the government through its Central Bank (BNM) would use this information to implement monetary and fiscal policies that help to control or minimize the impact of these factors on the stock market. Thus, policymakers and BNM could use monetary and fiscal policies to control money supply in order to achieve monetary and financial stability, because money supply indirectly influence in the appreciation or depreciation of the currency as well as in the rise or decline of inflation. Thus, having stable financial and monetary conditions would create favorable stock market returns, promoting economic growth in the country.

Moreover, a good understanding of the impact of crude oil price movements on stock market returns could help policymakers to regulate the stock market more effectively. Since the GFC of 2008, the KLCI dropped in value for the first time in 2014 due to low oil prices. This proves that understanding the movements of oil prices towards stock market would help to prevent or at least reduce the effect of oil price fluctuations in the Malaysia stock market.

Academicians and researchers may use this research paper to get more ideas between the macroeconomic variables and stock market performance in Malaysia. This information would help them in getting more familiarized with the stock market environment in Malaysia, and also assist them in conducting future research. Furthermore, based on the findings, investors, policymakers and the government should not consider much of foreign direct investment when making their decisions, since results show that this variable does not influence the variation of returns in the Malaysian stock market.

5.7. Conclusion

In a nutshell, this paper has investigated the impact of selected macroeconomic variables towards the stock market performance in Malaysia. Results have found that e-change rate and inflation have a negative effect towards the stock market, whereas crude oil prices and foreign direct investment are found to be positively correlated. Furthermore, e-change rate, inflation and crude oil prices turn to have a significant influence in the variation of Malaysian stock market returns, while foreign direct investment are found to have no impact in the stock market performance.. Moreover, this chapter has provided the implications the results obtained. Hence, findings from this study would provide useful information for a wide range of people such as investors, policymakers, government, academicians and researchers for different purposes. In short, the main objective of this paper which was to investigate the impact of e-change rate, inflation, crude oil price and foreign direct investment towards Malaysia stock market returns has been achieved.

Chapter 6: Limitations, Summary and Recommendations

6.0. Introduction

The last chapter of this research paper presents the limitations of this study, the summary of the overall content presented in the five previous chapters and also some recommendations for future researches.

6.1. Limitations

The current paper has also faced some limitations that people should be aware of. The first limitation is regarding the sample size or number of observations included in the research. The study was supposed to initially include 120 observations for each variable for the last 10 years (2007-2016). However, for Malaysia there was still no FDI data available for the year of 2016, which led to analyze only 108 observations for each variable, from the period of 2007 to 2015. Moreover, the research consists of monthly time series data for all the variables, but FDI data is only available in yearly and quarterly basis. Therefore, the data for this variable had to be converted into monthly basis, which can somehow reduce its accuracy.

Secondly, this research is based on time series data. Time series data such as GDP, e-change rate, money supply, stock inde- and others tend to face different problems such as heteroskedasticity, autocorrelation, multicollinearity, model specification and normality. When these problems e-ist in the model, it is a clear violation of the classical assumptions of OLS regression model. Consequently, the findings of the research are found to be biased. The model in this study faced normality problem which means that t-statistic and F-statistic would provide biased results. Nevertheless, OLS regression was still a consistent model to prove our hypotheses, since the other assumptions are all consistent with the classical assumptions of OLS regression model (Wooldridge, 2015).

Moreover, the paper selected four independent variables namely e-change rate, inflation, crude oil price and foreign direct investment to analyze their relationship with the Malaysian stock market (KLCI). However, the inclusion of some other variables such as money supply, interest rate, unemployment rate would certainly produce different results. Consequently, the inclusion of the four selected macroeconomic variables and the omission of other relevant variables from this study may still provide a biased result. In addition, KLCI represents only the 30 largest companies in Malaysia. Thus, findings from this paper may not be applicable for small and medium enterprises.

Last but not least, the research only investigates the Malaysian stock market and all the data use is based on Malaysia historical data. Hence, since the investigation is fully focused on Malaysia stock market, the findings are only useful for Malaysian policymakers and investors who wish to invest in the local stock market. Therefore, countries like Singapore, Indonesia, Brunei, Myanmar, Philippines, Laos, Cambodia, Thailand and Vietnam, in spite of falling under the same geographical area, they are not encouraged to apply the findings from this study due to different backgrounds, cultures, political factors and unique investment environment and economic policies.

6.2. Summary of the Study

The current research paper has analyzed the impact of four macroeconomic variables on the Malaysian stock market returns. The four macroeconomic variables consisted of e-change rate, inflation which was pro-ied by CPI, crude oil price and foreign direct investment. The Malaysian stock market was pro-ied by KLCI, which is the representative of the Malaysian stock market inde- since 1986, comprising the 30 largest companies in Malaysia. The research was carried out using monthly historical data from January 2007 to December 2015, collected from Thomson DataStream. Thus, each variable consisted of 108 observations in total.

Different diagnostic tests were performed including multicollinearity, heteroscedasticity, autocorrelation, model specification, normality as well as unit root tests using the econometric software E-views 9. These tests ensure that the results obtained are not bias, and thus they are reliable enough to be used for testing the hypotheses.

Results have shown that, in the short run e-change rate, and inflation are negatively correlated with the Malaysian stock market returns, and crude oil price together with FDI have a positive correlation the Malaysian stock market returns. Moreover, in the short run none of the four selected macroeconomic variables significantly affects the performance of the stock market in Malaysia. However, in the long run e-change rate, inflation and crude oil price e-cept FDI can be used to e-plain the variation of stock returns in Malaysia market. In addition, results suggest that all four variables together have a low forecast impact, which would be e-plained by the e-clusion of other important variables from the study.

6.3. Recommendations for Future Research

Future researchers are suggested to use individual company performance or industrial performance instead of using the KLCI as the pro-y variables to determine the relationship between the stock market performance and macroeconomic variables. Therefore, the research would be based on the weightage of each industry, instead of KLCI. The reason behind is because the KLCI only captures the 30 largest company in Malaysia regardless of their type, industry and nature. By analyzing the stock market using industrial performance may give more accurate and reliable results because more industries and companies can be included in the research, including small and medium enterprises which do not fit in the total KLCI weightage.

Moreover, researchers are encouraged to apply more relevant models that can better e-plain the effect of macroeconomic variables on stock market, in order to overcome the limitations of OLS regression model. According to Muneer, Butt & Rehman (2011) the model that can better capture the presence of time-varying stock price movement is the Generalized Auto Regressive Conditional Heteroscedasticity (GARCH). Moreover, Kasman et al. (2011) has argued that replacing OLS regression model with GARCH model can provide more efficient coefficients.

Lastly, future researchers can include larger sample size in their investigations to test the consistency of these results. This research included 108 monthly observations for each variable. Therefore, the sample size can be increased by increasing the frequency of the data such as using weekly or daily data. Since the stock market performance fluctuates a lot, increasing the sample size would enable researchers to obtain more concise results and also increase the reliability of the study.

7.0. References

Acheapong, I. K., & Wiafe, E. A. (2013). Foreign Direct Investment and Stock Market Development: Evidence from Ghana. International Journal of Finance and Policy Analysis, 5 (1). Pg.1-2.

Adusel, M. (2014). The Inflation-Stock Market Returns Ne-us: Evidence from the Ghana Stock E-change. Journal of Economics and International Finance, 6(2). Pg.38-46.

Agrawal, G., Kumar Srivastav, A. and Srivastava, A. (2010). A Study of E-change Rates Movement and Stock Market Volatility. International Journal of Business and Management, 5 (12). Pg. 62-63.

Ali, I. (2014). Impact of Foreign Direct Investment on Volatility of Stock Market: An Evidence from Pakistani Market. IOSR Journal of Business and Management, 16(1). Pg.77-80.

Al-Zoubi,M., & Al-Sharkas, A. A. (2011). Stock Prices and Inflation: Evidence from Jordan, Saudi Arabia, Kuwait, and Morocco. Paper presented at the Conference of Politics and Economic Development, Turkey.

Amanah Mutual Berhad (2015). Stock Market Highlights. AMB Mutual. Retrieved from: http://www.ambmutual.com.my/v2_/pdf/WEEKLY- INSIGHTS/2015/Jan/Week%20Ended%20January%202.pdf. [Accessed: 23rd May 2017].

Angabini,A., & Wasiuzzaman,S. (2011). Garch Models and the Financial Crisis: A study of the Malaysian Stock Market. The International Journal of Applied Economics and Finance, 5(3). Pg.226-236.

Arcabic, V., Globan, T. and Raguz, I. (2013). The Relationship between the Stock Market and Foreign Direct Investment in Croatia: Evidence from VAR and Co-integration Analysis. Financial Theory and Practice, 37(1). Pg.109-126.

Arouri,M.E.H. & Fouquau,J. (2011). On the Short-Term Influence of Oil Price Changes on Stock Markets in GCC Countries: Linear and Nonlinear Analyses. Journal of Economics Bulletin, 29(2). Pg. 795-804.

Asmy, M., Rohilina, W., Hassama, A., & Fouad, M. (2009). Effects of macroeconomic variables on stock prices in Malaysia: An approach of error correction model.

Asteriou, D. & Hall, S. G. (2014). Applied econometrics. London: Palgrave, Macmillan education.

Azizan, N. A. & Sorooshian, S., (2014). Stock Market Performance and Modern Portfolio Theory: Case on Malaysian Stock Market and Asian Indices. Wseas Transaction on Business and Economics, 111.

Basu, S. (1977). Investment Perfomance of Common Stocks in Relation to Their Price-Earnings Ratios: A Test of the Efficient Market Hypothesis. Journal of Finance,32(3). Pg.663-682.

Bello, Z. (2013). The Association between E-change Rates and Stock Returns. Investment Management and Financial Innovations, 10(3).

Bryman, A., & Bell, E., (2007). Business Research Methods. Ed. 2nd. O-ford: O-ford University Press.

Cakan, E & Ejara, D. (2013). On the Relationship between E-change Rates and Stock Prices: Evidence from Emerging Markets. International Research Journal of Finance and Economics. (111). Pg.115-124.

Chiang, T.C., Yang, S. Y. & Wang, T. S. (2000). Stock Return and E-change Rate Risk: Evidence from Asian Stock Markets Based on a Bivariate GARCH Model. International Journal of Business, 5.

Chinzara, Z. (2011). Macroeconomic Uncertainty and Conditional Stock Market 61 Volatility in South Africa. South African Journal of Economics, 79 (1).

Chong, F. & Puah, C. H. (2009). The Malaysian IPO Market: Volume, Initial Returns and Economic Conditions. Pg.1-11.

DeFina, R.H., (1991). Does Inflation Depress the Stock Market. Federal Reserve Bank of Philadelphia Business Review, Nov., 3-12.

Dickey, D. A & Fuller, W. A. (1976). Distribution of The Estimates for Autoregressive Time Series With A Unit Root. Journal of the American Statistical Association, 74. Pg.427- 431.

Din, S. M. & Mubasher, H.M. (2013). Macroeconomic Variables on Stock Market Interactions: The Indian E-perience. Advances in Management, 6(8).

Durbin, J., & Watson, G. S. (1950). Testing for serial correlation in least squares regression: I. Biometrika, 37(3/4), 409-428.

Eagle, R. F. (1982). Autoregressive Conditional Heteroskedasticity with Estimates of the

Variance of U.K. Inflation. Econometrics, 50. Pg.987-1008.

Eita, H. J. (2012). Modelling Macroeconomic Determinants of Stock Market Prices: Evidence from Namibia. Journal of Applied Business Research, 28(5), 167-177.

El-Nader, H. and Alraimony, A. (2012). The Impact of Macroeconomic Factors on Amman Stock Market Returns. International Journal of Economics and Finance, 4(12). Pg. 202- 2013.

Fama, E. F. (1969). Efficient Capital Markets: A Review of Theory and Empirical Work. The Journal of Finance, 25(2). Pg.383-417.

Fang, C. and You, S. (2014). The Impact of Oil Price Shocks on the Large Emerging Countries' Stock Prices: Evidence from China, India and Russia. International Review of Economics & Finance, 29. Pg.330-338.

Fang, W.S. and Miller S. M. (2002). Currency Depreciation and Korean Stock Market Performance During the Asian Financial Crisis. Working Paper, University of Connecticut.

Farmer, R. (2015). The Stock Market Crash Really Did Cause the Great Recession. O-ford Bulletin of Economics and Statistics, 77(5). Pg.617-633.

Farrukh Ahmed, S., Islam, K. and Khan, M. (2015). Relationship between Inflation and Stock Market Returns: Evidence from Bangladesh. Daffodil International University Journal of Business and Economics, 9(1). Pg.1-9.

Fatimah, N. & Shamim, A. S. (2012), “Empirical Evidence of Fisher Effect in Pakistan,” World Applied Sciences Journal, 18(6), 770-773.

Geetha, C. Mohidin, R. Chandran, V. V. & Chong, V. (2011). The Relationship between Inflation and Stock Market: Evidence from Malaysia, United States and China. International Journal of Economics and Management Sciences, 1(2). Pg.1-16.

Granger, C. W. (1969). Investigating causal relations by econometric models and cross-spectral methods. Journal of the Econometric Society. Pg.424-438.

Gujarati, D. N., & Porter, D. C. (2009). Basic econometrics. 1st edition. McGraw-Hill.REF ID, 36.

Gupta, R. & Inglesi-Lotz, R. (2012). Macro Shocks and Real US Stock Prices with Special Focus on the "Great Recession". Applied Econometrics and International Development, 12 (2). Pg.123-136.

Harasheh, M., & Abu-Libdeh, H. (2011). Testing for Correlation and Causality Relationship between Stock Prices and Macroeconomic Variables: The Case of Palestine Securities E-change. International Review of Business Research Papers, 7(5). Pg.141-154.

Hossenidoust, E. et al. (2013). Volatility Spillovers across Commodity and Stock Markets among ASEAN Countries. Prosiding Persidangan Kebangsaan Ekonomi Malaysia, 3. Pg.1401- 1412.

Hsing,Y. (2011). Macroeconomic Determinants of the Stock Market Inde- and Policy Implications: The Case of a Central European Country . Eurasian Journal of Business and Economics, 4(7). Pg.1-11.

Humpe, A. & Macmillan, P.D. (2007). Can Macroeconomic Variables E-plain Long-Term Stock Market Movements? A Comparison of the US and Japan. Working series paper. Centre for Research into Industry, Enterprise, Finance and the Firm.

Idenyi, O. S. Ifeyinwa, A. C. Obinna, N. J. and Promise, E. A. (2016). Impact of Foreign Direct Investment on Stock Market Growth in Nigeria. Asian Research Journal of Arts & Social Sciences, 1(2).

Irshad Younas, Z., Farooq, U. and Nasir, A. (2016). E-change rate and stock market volatility: A case of Pakistan. Journal of Business Administration and Management Sciences Research, 2 (8). Pg. 187-190.

Jafarian, A., & Safari, M. (2015). Impact of Oil Price Fluctuations on Returns of Different Sectors of Malaysian Stock Market. Journal of Modern Accounting and Auditing, 11(3). Pg.159-169.

Jarque, C. M., & Bera, A. K. (1987). A Test for Normality of Observations and Regression Residuals. International Statistical Review, 55(2).

Jawaid, S. T. & Haq, A. U. (2012). Effects of Interest Rate, E-change Rate and Their Volatilities on Stock Prices: Evidence from Banking Industry of Pakistan. Theoretical and Applied Economics, 19(8).

Kasman, S., Vardar, G., & Tunç, G. (2011). The Impact of Interest Rate and E-change Rate Volatility on Banks’ Stock Returns and Volatility: Evidence from Turkey. Economic Modelling, 28.

Khan, Z., Khan, S., Rukh, L. and Rehman, W., U. (2016). Impact of Interest Rate, E-change Rate and Inflation on Stock Market Returns of KSE 100 Inde-. International Journal of Economic Research, 3(5).

Koch, C. (1996). The Arbitrage Pricing Theory as an Approach to Capital Asset Valuation. GRIN Verlag.

Kutty, G. (2010). The Relationship between E-change Rates and Stock Prices: The Case of Me-ico. North American Journal of Finance and Banking Research, 4(4). Pg.1-12.

Kuwornu, J.K.M. (2012). Effect of Macroeconomic Variables on the Ghanaian Stock Market Returns: A Cointegration Analysis. Agris Online Papers in Economics and Informatics, 4(2).

Le, T. H., & Chang, Y. (2011). The Impact of Oil Price Fluctuations on Stock Markets in Developed and Emerging Economies. Nanyang Technological University, School of Humanities and Social Sciences, Economic Growth Centre Working Paper No. 2011/03. Lim, C. & Sek, S. (2013). Comparing the Performances of GARCH-type Models in Capturing the Stock Market Volatility in Malaysia. Procedia Economics and Finance, 5. Pg.478- 487.

Lintner, J. (1965). The Valuation of Risky Assets and the Selection of Risky Investment in Stock Portfolios and Capital Budgets. The Review of Economics and Statistics, 47(1), 13-37.

Long, J. S., & Laurie, H. E. (1998). Correcting for Heteroscedasticity with Heteroscedasticity Consistent Standard Errors in the Linear Regression Model. Retrieved from: http://www.indiana.edu/~jslsoc/files_research/testing_tests/hccm/00TAS.pdf. Maghyereh, A. (2004). Oil Price Shocks and Emerging Stock Markets: A generalized VAR Approach. International Journal of Applied Econometrics and Quantitative Studies, 1(2). Masoud, M. H. N. (2013). The Impact of Stock Market Performance upon Economic Growth. International Journal of Economics and Financial Issues, 3(4). Pg. 789.

Maysami, R. C. & Koh, T. S. (2000). A Vector Error Correction Model of the Singapore Stock Market. International Review of Economics & Finance, 9(1). Pg.79-96.

Melnyk, L., Kubatko, O. and Pysarenko, S. (2014). The impact of Foreign Direct Investment on Economic Growth: Case of Post Communism Transition Economies. Problems and Perspectives in Management, 12 (1). Pg.17-24.

Merton, R. (2010). An Intertemporal Capital Asset Pricing Model. Econometrica: Journal of the Economic Society, 41(5), p.867.

Mlambo, C., Maredza, A. & Sibanda, K. (2013). Effects of E-change Rate Volatility on the Stock Market: A Case Study of South Africa. Mediterranean Journal of Social Sciences, 4 (4). Pg.562.

Mohanty, S. K., Nandha, M., Turkistani, A. Q., Alaitani, M. Y., (2011). Oil Price Movements and Stock Market Returns: Evidence from Gulf Coorporation (GCC) Countries. Global Finance Journal, 22. Pg.42-55.

Mohanty, S. K., Nandha, M., Turkistani, A. Q., & Alaitani, M. Y. (2011). Oil Price Movements and Stock Market Returns: Evidence from Gulf Coorporation (GCC) Countries. Global Finance Journal, 22.

Mollick, A. & Assefa, T. (2013). U.S. Stock Returns and Oil Prices: The Tale from Daily Data and the 2008-2009 Financial Crisis. Energy Economics, 36. Pg.1-18.

Mousa, N. S., et al. (2012). The Relationship between Inflation and Stock Price: A Case from Jordan. IJAARS, 10 (1). Pg.46-52.

Muhammad, N. and Rasheed, A. (2012). Stock Prices and E-change Rates: Are They Related? Evidence from South Asian Countries. The Pakistan Development Review, 535-550.

Muneer, S., Butt, B. Z., & Rehman, K. (2011). A Multifactor Model of Banking Industry Stock Returns: An Emerging Market Perspective. Information Management and Business review, 2(6).

Musa, S. U., & Ibrahim, M. (2014). Stock Market Development, Foreign Direct Investment and Macroeconomic Stability: Evidence from Nigeria. Research Journal of Finance and Accounting, 5 (18). Pg.258-263.

Najaf, R. (2016). Impact of International Oil Prices on the Stock E-change of Malaysia and Turkey. Journal of Accounting & Marketing, 5(4).

Najaf, R. and Najaf, K. (2016). A Study of E-change Rates Movement and Stock Market Volatility. Asian Journal of Management, Engineering & Computer Sciences (AJMECS), 1(1). Pg.32.

Nordin, S. & Nordin, N., (2016). The Impact of Capital Market on Economic Growth: A Malaysian Outlook. International Journal of Economics and Financial Issues ISSN: 2146-4138. 6(7). Pg. 259-265.

of Malaysian Stock Market. Journal of Modern Accounting and Auditing. 11(3).

Olugbenga, A. A. & Grace, O. O. (2015). Impact of Foreign Direct Investment on Nigerian Capital Market Development. International Journal of Academic Research in Accounting, Finance and Management Sciences, 5(1). Pg.103-108.

Onyinyechi, O. C. & Ekwe, M. C. (2016). Impact of Foreign Direct Investment (FDI) on the Stock Market Performances In Nigeria (1985-2014). Applied Finance and Accounting, 3(1).

Osho, A. (2014). The Role of Stock Market on Nigeria’s Economic Development. International Journal of Scientific and Research Publications, 4(4). Pg.1.

Ouma, W. N. & Muriu, P. (2014). The Impact of Macroeconomic Variables on Stock Market Returns in Kenya. International Journal of Business and Commerce, 3(11).

Phuyal, N. (2016). Can Macroeconomic Variables E-plain Long Term Stock Market Movements? A Study of Nepali Capital Market. Journal of Business and Management Research, 1(1).

Pönkä, H. (2016). Real oil prices and the international sign predictability of stock returns. Finance Research Letters, 17. Pg.79-87.

Poole, W. (2010). Causes and Consequences of the Financial Crisis of 2007-2009. Harvard Journal of Law & Public Policy, 33. Pg.421-441.

Qamri, G. M., Haq, M. A. U & Akram, F (2015). The Impact of Inflation on Stock Prices: Evidence from Pakistan. Microeconomics and Macroeconomics, 3(4).

Reboredo, J. and Rivera-Castro, M. (2014). Wavelet-based evidence of the impact of oil prices on stock returns. International Review of Economics & Finance, 29. Pg.145-176.

Roll, R. and Ross, S. (1976). The Arbitrage Pricing Theory Approach to Strategic Portfolio Planning. Financial Analysts Journal.

Sari, R. & Soytes, U. (2006). The Relationship between Stock Returns, Crude Oil Prices, Interest Rates and Output: Evidence from a Developing Economy. The Empirical Economics Letters, 5(4).

Saunders, M., Philip, L. & Thornhill, A.,(2012). Research Methods for Business Students. Ed. 6th. Harlow: Financial Times Prentice Hall.

Sekar, P. (2010). Diagnostic Checking of Time Series Models. Indian Journal of Science and Technology, 3(9).

Shahbaz, M., Lean, H. and Kalim, R. (2013). The Impact of Foreign Direct Investment on Stock Market Development: Evidence from Pakistan. Economic Research-Ekonomska Istra ž ivanja, 26(1). Pg.17-32.

Sharpe, W. F. (1964). Capital Assets Prices: A Theory of Market Equilibrium under Conditions of Risk. T he Journal of Finance, 19(3), 425-442.

Soenen, L. C. and Hennigar, E. S. (1988). an Analysis of E-change Rates and Stock Prices: The US E-perience between 1980 and 1986. Akron Business and Economic Review Winter, 19.

The Economic Times, (2016). Volatility. [Online]. Available at: http://economictimes.indiatimes.com/definition/volatility. [accessed: 23rd September 2016].

Tsai, C. (2015). How do U.S. Stock Returns Respond Differently to Oil Price Shocks Pre-crisis, Within the Financial Crisis, and Post-crisis? Energy Economics, 50. Pg.47-62.

Tu, Y. and Tan, X. (2012). Technology spillovers of FDI in ASEAN Sourcing from Local and Abroad. China Finance Review International, 2(1). Pg .78-94.

Uwubanmwen, A. & L. Eghosa, I. (2015). Inflation Rate and Stock Returns: Evidence from the Nigerian Stock Market. International Journal of Business and Social Science, 6(11). Pg.155-165.

Vejzagic, M. & Zarafat, H. (2013). Relationship between Macroeconomic Variables and Stock Market Inde-: Co-integration Evidence from FTSE Bursa Malaysia Hijrah Shariah Inde-. Asian Journal of Management Sciences & Education, 2(4), 94-108.

White, H. (1980). A Heteroskedasticity-Consistent Covariance Matri- Estimator and a Direct Test for Heteroskedasticity. The Econometric Society, 48(4). Pg.817-38.

Wooldridge, J. M. (2015). Introductory Econometrics: A Modern Approach. 6th Edition.

Cengage Learning.

Yau, H. Y. & Nieh, C. C. (2006). Interrelationships among Stock Prices of Taiwan and Japan and NTD/Yen E-change Rate. Journal of Asian Economics, 17.

Yau, H., & Nieh, C. (2009). Testing for Cointegration with Threshold Effect between Stock Prices and E-change Rates in Japan and Taiwan. Japan and the World Economy, 21. Zabarankin, M. Pavlikov, K. & Uryasev, S. (2013). Capital Asset Pricing Model (CAPM) with Drawdown Measure. European Journal of Operational Research, 234.

Zhang, C. G., & Chen, X. Q. (2011). The Impact of Global Oil Price Shocks on China’s Stock Returns: Evidence from the ARJI (-ht)-EGARCH model. Energy, 36(11). Pg.6627-6633.

Appendices

Appendi- 1: Unit Root Test

- Augmented Dickey Fuller Test at Level Null Hypothesis: LNKLCI has a unit root

E-ogenous: Constant

Lag Length: 0 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNER has a unit root E-ogenous: Constant

Lag Length: 0 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNFDI has a unit root E-ogenous: Constant

Lag Length: 3 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNCOP has a unit root E-ogenous: Constant

Lag Length: 1 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNCPI has a unit root E-ogenous: Constant

Lag Length: 1 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNKLCI has a unit root

E-ogenous: Constant, Linear Trend

Lag Length: 1 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNER has a unit root E-ogenous: Constant, Linear Trend

Lag Length: 0 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNFDI has a unit root E-ogenous: Constant, Linear Trend

Lag Length: 3 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNCPI has a unit root E-ogenous: Constant, Linear Trend

Lag Length: 1 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNCOP has a unit root E-ogenous: Constant, Linear Trend

Lag Length: 1 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

- Augmented Dickey Fuller at First Difference

Null Hypothesis: D(LNKLCI) has a unit root E-ogenous: Constant

Lag Length: 0 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNER) has a unit root E-ogenous: Constant

Lag Length: 0 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNFDI) has a unit root E-ogenous: Constant

Lag Length: 2 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNCPI) has a unit root E-ogenous: Constant

Lag Length: 0 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNCOP) has a unit root E-ogenous: Constant

Lag Length: 0 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNKLCI) has a unit root

E-ogenous: Constant, Linear Trend

Lag Length: 0 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNER) has a unit root E-ogenous: Constant, Linear Trend

Lag Length: 0 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNFDI) has a unit root E-ogenous: Constant, Linear Trend

Lag Length: 2 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNCPI) has a unit root E-ogenous: Constant, Linear Trend

Lag Length: 0 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNCOP) has a unit root E-ogenous: Constant, Linear Trend

Lag Length: 0 (Automatic - based on SIC, ma-lag=12)

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

- Phillips-Peron Test at Level

Null Hypothesis: LNKLCI has a unit root E-ogenous: Constant

Bandwidth: 5 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNER has a unit root E-ogenous: Constant

Bandwidth: 3 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNFDI has a unit root E-ogenous: Constant

Bandwidth: 6 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNCPI has a unit root E-ogenous: Constant

Bandwidth: 1 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNCOP has a unit root E-ogenous: Constant

Bandwidth: 4 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNKLCI has a unit root

E-ogenous: Constant, Linear Trend

Bandwidth: 6 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNER has a unit root E-ogenous: Constant, Linear Trend

Bandwidth: 8 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNFDI has a unit root E-ogenous: Constant, Linear Trend

Bandwidth: 6 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNCPI has a unit root E-ogenous: Constant, Linear Trend

Bandwidth: 1 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: LNCOP has a unit root E-ogenous: Constant, Linear Trend

Bandwidth: 4 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNER) has a unit root E-ogenous: Constant

Bandwidth: 4 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNFDI) has a unit root E-ogenous: Constant

Bandwidth: 6 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNCPI) has a unit root E-ogenous: Constant

Bandwidth: 4 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNCOP) has a unit root E-ogenous: Constant

Bandwidth: 2 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNKLCI) has a unit root

E-ogenous: Constant, Linear Trend

Bandwidth: 4 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNER) has a unit root E-ogenous: Constant, Linear Trend

Bandwidth: 10 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNFDI) has a unit root E-ogenous: Constant, Linear Trend

Bandwidth: 6 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNCPI) has a unit root E-ogenous: Constant, Linear Trend

Bandwidth: 4 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Null Hypothesis: D(LNCOP) has a unit root E-ogenous: Constant, Linear Trend

Bandwidth: 2 (Newey-West automatic) using Bartlett kernel

Abbildung in dieser Leseprobe nicht enthalten

*MacKinnon (1996) one-sided p-values.

Appendi- 2: Heteroskedasticity Test

Heteroskedasticity Test: Breusch-Pagan-Godfrey

Abbildung in dieser Leseprobe nicht enthalten

Test Equation:

Dependent Variable: RESID^2 Method: Least Squares

Date: 08/20/17 Time: 16:39 Sample: 2007M02 2015M12 Included observations: 107

Abbildung in dieser Leseprobe nicht enthalten

Appendi- 3: Autocorrelation

Breusch-Godfrey Serial Correlation LM Test:

Abbildung in dieser Leseprobe nicht enthalten

Test Equation:

Dependent Variable: RESID Method: Least Squares Date: 08/20/17 Time: 16:46 Sample: 2007M02 2015M12 Included observations: 107

Presample missing value lagged residuals set to zero.

Abbildung in dieser Leseprobe nicht enthalten

Appendi- 5: Model Specification

Equation: UNTITLED

Specification: DLNKLCI C DLNCOP DLNCPI DLNER DLNFDI Omitted Variables: Squares of fitted values

Abbildung in dieser Leseprobe nicht enthalten

Unrestricted Test Equation:

Dependent Variable: DLNKLCI Method: Least Squares

Date: 08/20/17 Time: 18:04 Sample: 2007M02 2015M12 Included observations: 107

Abbildung in dieser Leseprobe nicht enthalten

Appendi- 6: Granger Causality Test

Pairwise Granger Causality Tests

Date: 08/21/17 Time: 22:32 Sample: 2007M01 2015M12 Lags: 2

Abbildung in dieser Leseprobe nicht enthalten