The Impact of Macroeconomic Variables on Al-Quds Index: Empirical Evidence from Palestine

International Journal of Financial Economics Vol. 3, No. 5, 2014, 228-241 The Impact of Macroeconomic Variables on Al-Quds Index: Empirical Evidence from Palestine Ibrahim Awad 1, Mohanad Rezeq 2, Abdelrahman Aliwisat 3 Abstract This study aimed at investigating the impact of macroeconomic variables, particularly GDP, CPI, unemployment, exchange rate, and interest rate on the Al-Quds Index (AQI). Also, the causality relationships between these variables were examined. Toward that end, the study used quarterly data observations over the period from 2002 to 2012. Accordingly, the study found that AQI isn t significantly linked with macroeconomic variables. Granger causality indicates that AQI does not Granger causes the macroeconomic variables that been tested, and the macroeconomic variables also do not Granger cause the AQI, so that the macroeconomic variables cannot explain the behavior of the AQI due to the inefficiency of the AQI. For policy makers, this study confirms the necessity of future researches so as to significantly evaluate the Palestine Exchange (PEX). Keywords: Granger Causality test, Macroeconomic Variables, Palestine Exchange, Regression Analysis, and Stationarity. JEL Classification: E44, E47, C01. 1. Introduction Financial markets play a central role in the allocation of capital resources. Investors in the stock market can expect which company will continue and they evaluate the financial position of a listed firm. If a corporation seems to have good prospects for future profitability, investors will bid up its stock price. The company s management will find it easy to issue new shares or borrow funds to finance research and development, build new production facilities, and expand its operations (Bodie et al., 2009). Through this simple definition of financial markets, we conclude that the financial market plays an important role so as to improve the economic development. With reference to, Buyuksalvarci (2010) Capital markets play an important role in the financial sector of each economy. An efficient capital market can promote economic growth and prosperity by stabilizing the financial sector and providing an important investment channel that contributes to attracting domestic and foreign capital. According to the Palestine Exchange (PEX) website, the PEX was established in 1995 to promote investment in Palestine. It became a public shareholding company in February 2010 responding to the principles of transparency and good governance. The PEX mainly strives to provide an enabling environment for trading that is characterized by equity, transparency and competence, serving and maintaining the interest of investors and it is very appealing in terms of market capitalization. Added to this, it passed with the minimum level of impact of the global financial crisis compared to other MENA Exchanges. As of today, there are 46 listed companies on PEX with a market capitalization of about $ 2.7 billion across five main economic sectors; (1) banking and financial services; (2) insurance; (3) investments; (4) industry; and (5) services. Most of the listed companies are profitable and trade in Jordanian Dinar, while 1 Faculty of Economics and Business, Al-Quds Unversity P. O. Box 51000, Jerusalem 2 Institute of Economics and Business, Al-Quds University, East Jerusalem, Palestine 3 Talal Abu-Ghazaleh & Co. International, Ramallah, Palestine 2014 Research Academy of Social Sciences http://www.rassweb.com 228

Q1 2002 Q3 2002 Q1 2003 Q3 2003 Q1 2004 Q3 2004 Q1 2005 Q3 2005 Q1 2006 Q3 2006 Q1 2007 Q3 2007 Q12008 Q3 2008 Q12009 Q3 2009 Q12010 Q3 2010 Q12011 Q3 2011 Q1 2012 Q3 2012 International Journal of Financial Economics others trade in US Dollars. Only stocks are currently traded on PEX, but there is potential and readiness to trade other securities in the future. In the course of this study we will cast light on the impact of macroeconomic variables especially GDP, CPI, unemployment, exchange rate, and interest rate on the Al-Quds Index (AQI). As many empirical studies over the world focused on the impact of macroeconomic variables on stock market indices (Abu-Libdeh and Harasheh, 2011), Hsing (2011), Hsing et al. (2012), Sohail and Hussain (2011), Gonzalo and Taamouti (2011), D.Gay,Jr, (2008), Binti Mohamed et al. (2011), McQueen and Roley (1993)). Considering that the AQI is the main index of PEX and is composed of 12 listed companies across all business sectors. The sample companies included in AQI are changed at the beginning of every year in order to include the most active companies in terms of value of traded stocks, the amount of traded stocks, total number of transactions, number of trading days, the stock turnover ratio, and market value of companies at the end of the previous year.there are also other five indices works in PEX : PEX General Index (composed of all listed companies), Banking and Financial Services Index, Industry Index, Insurance Index, and Investment Index ( PEX website). 1200 1000 800 600 400 200 0 Al-Quds Index (AQI) values Figure 1: Quarterly values of AQI from 2002 to 2012. The overall obective of this study is to provide policy makers and investors with important and practical information on macroeconomic variables that may have or have not impact on the Al Quds index. Specifically, the study will examine whether or not the macroeconomic variables used in this study affect the value of Al Quds index. 2. Previous Research Many empirical studies were tried to find how macroeconomic variables can affect stock market index in developed and underdeveloped countries. In addition, some of these studies focus on the relationship between these macroeconomic variables and stock market indices. Abu-Libdeh and Harasheh (2011) aims at investigating the correlation and causality relationships between stock prices in Palestine and some macroeconomic variables by using this model( QI = α + β1 GDP + β2 INF + β3 Ex + β4 LIBOR + β5 BOT + е) Where α is the vertical intercept, β is the regression coefficients and e is the error term, and then they found by using the regression analysis that AQI have positive and significant relationship with GDP, Inflation rate, Exchange rate, and LIBOR but negative relationship with BOT but not significant. Added to this Hisng (2011) examined the relationship between the South African stock market index and selected macroeconomic variables and he found that more real GDP growth, a lower ratio of the government deficit to GDP, a higher ratio of M3 to GDP, a lower domestic real interest rate, depreciation of the rand, a lower inflation rate, a higher U.S. stock price, or a lower U.S. government bond yield would help the South African stock market. 229

I. Awad et al. On the other hand, Hsing et al. (2012) examined the macroeconomic factors that are expected to influence the Argentine stock market index. The paper shows that the Argentine stock market index has a positive relationship with real GDP, the M2/GDP ratio, the peso/usd exchange rate and the U.S. stock market index and a negative relationship with the money market rate,the ratio of government spending to GDP and the inflation rate. All independent variables are significant at the 1% level. Further that, Hsing (2011) examined the effects of macroeconomic variables on the stock market: the case of the Czech Republic. More real GDP, a lower government borrowing/gdp ratio, a lower real interest rate or expected inflation rate, a higher US or German stock market index, or a lower euro area government bond yield would increase the Czech stock market index, at the same time he examined the relationship between Hungary s stock market index and relevant macroeconomic variables, Hsing (2011). All the coefficients are significant at the 1% level. The stock market index has a positive relationship with real GDP, the government debt/gdp ratio, the nominal effective exchange rate and the German stock market indexes, a negative relationship with the real Treasury bill rate, the expected inflation rate and the government bond yield in the euro area, and a quadratic relationship with real M2 money supply. Additionally, some studies showed the impact of macroeconomic variables on Asian stock market indices likes, Sohail and Hussain (2011) aims to investigate the long run, and short run dynamics relationships between KSE100 index and five macroeconomic variables in Pakistan. The results revealed that in the long run, there was a positive impact of inflation, GDP growth, and exchange rate on KSE100 index,, while money supply and three months treasury bills rate had negative impact on the stock returns. Also Binti Mohamed et al. (2011) examined the correlation between Malaysian Sectoral Indices and Macroeconomic variables. They show that there is a negative relationship between CPI with consumer product and industrial product index in Bursa Malaysia. And there is a negative relationship between interest rate (BLR) with consumer product and industrial product index in Bursa Malaysia. Results also show that M2 has a positive relationship with consumer product and industrial product index in Bursa Malaysia which means that all variables have significant relationship with the stock market indices. The anticipated unemployment rate has a strong impact on stock prices, that means an increase in the anticipated unemployment rate leads to an increase in the stock market price (return).because if the unemployment rate is high, the Fed decreases the interest rate which in turn increases the stock market prices Gonzalo and Taamouti (2011). Also McQueen and Roley (1993) found a strong relationship between stock prices and macroeconomic news, such as news about inflation, industrial production, and the unemployment rate based on their own definition of business conditions. However their purpose was to demonstrate the dependency of stock price responses to all macroeconomic news. Hypotheses Hypothesis 1: Empirical work confirms that there is a relationship between price index and macroeconomic variables, so we expect H 0 : Macroeconomic variables aren t related to AQI H A : Macroeconomic variables are related to AQI Hypothesis 2: Granger causality technique is used to test whether or not available lagged information on macroeconomic variables has any significant effect on Al-Quds index. If not, macroeconomic variables don t Granger cause Al-quds Index. This test considers whether or not macroeconomic variables do Granger cause Al-Quds Index. Let macroeconomic variables to be X and Al-Quds Index to be Y. If a variable X is found to be helpful for predicting variable Y, then a time series X Granger causes a time series Y and Y can be forecasted by values of X. Therefore, we expect a causal relationship between Al-Quds index and Macroeconomic variables. To capture causality between Al-Quds index and Macroeconomic variables, vector autoregression (VAR) models are used under two cases: 230

International Journal of Financial Economics H 1 : xt i xt i p 1 y i 1 1 H 0 : Ʃβ i = 0 H A : Ʃβ i 0 If the null hypothesis is reected, macroeconomic variables are said to Granger cause AQI. H 0 : Ʃθ i = 0 H A : Ʃθ i 0 H 2 : yt i yt i p i 1 1 q t t x, t 2 x q y, t If the null hypothesis is reected, Al-Quds Index is said to Granger cause macroeconomic variables. In addition, if the null hypotheses are reected under the two cases, there is mutual causal relationship between macroeconomic variables and Al-Quds Index, otherwise there is no feedback relationship among these variables. 3. Methodology and Data In order to meet the study obectives, the study undertakes quantitative analysis of time series data. Toward that end, econometric techniques of unit root tests and causality tests were used, which are to investigate whether or not macroeconomic variables do cause Al-Quds Index. Regression analysis was also used to examine the impact of macroeconomic variables on Al-Quds Index. Data This study covers this period for analysis from 2002 to 2012. In particular, quarterly data was used to discover the association between the macroeconomic variables and Al-Quds index in PEX. All the data were collected from three main sources, the Palestinian Central Bureau of Statistics (PCBS), Bank of Israel, and Palestinian Exchange(PEX). Al-Quds Index is measured by stock price index with 2004 as the base year. GDP is represented by the real gross domestic product in millions at constant price. CPI is represented by the consumer pricing index with 2004 as the base year. UNEMP is represented by the rate of unemployment in Palestinian territory from year 2002 to the year 2012. EXR is measured by the NIS/USD exchange rate. An increase in the NIS/USD exchange rate means depreciation of the Israeli Shekel. The choice of the NIS/USD exchange rate is because the maor currency in the Palestinian Authority is the Israeli Shekel and the currency trading in the stock exchange of Palestine is the U.S. dollar and the Jordanian dinar. IR is taken from Bank of Israel Since the Palestinian economy is an economy that is complementary to the Israeli economy. Variables to be Investigated Gross Domestic Product (GDP) Gross Domestic Product (GDP) in the Palestinian territory found by economic activity for the quarters of the years 2000-2011 at constant prices, and it calculated from total items of economic activities (Agriculture and fishing, Mining and quarrying, Manufacturing, Electricity and water supply, Construction, Wholesale and retail trade, Transport, Storage and Communications, Financial intermediation, Real estate, renting and business services, Community, social and personal services, Hotels and restaurants, Education, Health and social work, Public administration and defense, Households with employed persons, FISIM, customs duties, VAT on imports, and net) PEX website. An empirical studies by Abu-Libdeh and Harasheh (2011), Hsing (2011), Hsing et al. (2012), and Sohail and Hussain (2011) conclude that GDP has positive effects on the stock market. 231

I. Awad et al. Consumer Price Index (CPI) Consumer Price Index is used as a proxy of inflation rate. Quarterly consumer price index (CPI) numbers calculated from average total items of expenditures, and the maor groups of expenditure are (Food and soft drinks, Al coholic Beverages and tobacco, Textiles, clothing and footwear, Housing, Furniture, household goods, Medical care, Transportation, Communications, Recreational, cultural goods & service, Education, Restaurants and cafes, and Miscellaneous goods and services ) PEX website. Many studies shows that have negative effect of CPI on the stock market, Yu Hsing (2011), Hsing et al. (2012), and Binti Mohamed et al. (2011), but Sohail and Hussain (2011) found that the effect of CPI on the market index has positive. Unemployment (UNEMP) The number of unemployed persons was calculated under the relaxed definition by adding to unemployed persons according to the ILO Standards, those persons outside labor force because they were frustrated. PEX website. Gnzalo and Taamouti (2011) found that the anticipated unemployment rate has positive relation with the stock market return in the short-run.and ustified this by the increase of unemployment rate lead the Fed decrease the interest rate which in turn increase the stock market prices. Exchange Rate (EXR) In this study quarterly (USD/NIS) exchange rate is employed as foreign exchange rate. Added to this, they obtained from Bank of Israel from 2000 to the second quarter of 2011. There are many studies examine the effect of exchange rate on the stock market index, like Hsing et al. (2012), Hsing (2011), Sohail and Hussain (2011), and D.Gay, Jr, (2008) show that Exchange Rate has positive effect on the market index. Interest Rate (IR) There are two kinds of interest rates are undertaken in the Palestinian Territory: (1) fixed interest is around 11% annually on the dollar, computed on a daily basis while charged on monthly basis; and (2) interest rate may depend on the London Interbank Offered Rate (LIBOR), and is calculated by adding a fixed rate (2%-3% annually) to the LIBOR. The interest is computed on a daily basis but charged monthly. In this study we tused the interest rate from bank of Israel to show if there have any impact of Israeli interest rate on AQI in PEX. Hsing (2011), Sohail and Hussain (2011), Gnzalo and Taamouti (2011), and Binti Mohamed et al. (2011) find that interest rate has negative effect on the stock market index. We choose this five independent variables in this study depended on four main factors. First, by reviewing previous studies we found that these variables have more influence than others, second, the nature of the Palestinian economy is an emerging economy with different conditions for other emerging economies in the world, third, do not have a central bank in Palestinian Territory that control of money supply in the market and control of the interest rate, final, other variables were tested and omitted due to the multicollinearity effect that make deficiencies in regression analysis, Haneen, Murad (2011). Model of the Study Two models have been used to test the relationships between macroeconomic variables and Al-Quds Index namely multiple regression and Granger Cuasality technique. Multiple Regression As this study examined the effects of macroeconomic variables on Al-Quds index, it is more feasible to use multiple regression model in this respect. This model was useful and suitable because the research focus lied in examining the contemporaneous relationships between stock returns and changes in macroeconomic variables. With reference to both theoretical and empirical literature reviewed, this study hypothesize the model between Al-Quds index and five macroeconomic variables, namely gross domestic product, consumer price 232

International Journal of Financial Economics index, unemployment rate, exchange rate, and interest rate. The hypothesized model is represented as follows: Y = f ( GDP, CPI, UNEMP, EXR, IR) In order to see whether the above macroeconomic factors could explain Al-Quds index returns, the multiple regression model is formed: AQI = β 0 + β 1 GDP + β 2 CPI + β 3 UNEMP + β 4 EXR + β 5 IR + ε, where; AQI= Quarterly AQI value; GDP = Quarterly gross domestic product; CPI = Quarterly consumer price index; UNEMP = Quarterly unemployment rate; EXR = Quarterly exchange rate; and IR = Quarterly interest rate. Where, β 0 is constant and the other β s are coefficients of the dependent variables while ε is the residual error of the regression. The four important parts of the output of regression are: Accuracy of R- squared, significance of F of the regression, reliability of coefficients, and residuals show no patterns. Thus, we will determine whether or not dependent variable is affected by independent variable(s). In order to test the goodness of fit for the model, adusted R-squared, Durbin Watson test, VIF, and T- test were tested and taken into account to ascertain the significance of the parameters at 5% level of significance. Granger Causality This test involves time series data to be stationary, so normality and stationarity tests have to be conducted for time series as an initial step to investigate causality between the al-quds Index and Macroeconomic variables. Stationarity Test (Unit Root Test): When the error term is a function of its past values, there is an autoregression, which is expressed as follows: E t = ρe t-1 + V t, which means that the error term at period t is a function of itself in the previous time period t-1 times ρ, which is the first order autocorrelation coefficient, and V denotes the white noise error term and it is random. If there is an autocorrelation in error term, the data aren t stationary. However, data should be stationary in order to make predictions and receiving reliable results. So, non-stationary data need to be converted into stationary data. To test stationarity, Augmented Dickey Fuller (ADF) and Philip Person (PP) are used. Khan et al., (2010), Ghaith and Awad (2011), Awad and Daraghma (2009), and Tudor (2011) used ADF test for higher order correlation by assuming that the series follows an autoregression error scheme. Also, they augmented the ADF approach controls for higher order correlation by adding lagged difference terms of the dependent variable. This test is used initially to find whether or not data points present stationary process. Therefore, we will use unit root tests by utilizing a Dickey Fuller test in order to test the stationarity of time series data. This equation indicates that current changes in stock price is a function of past realization, SP t = f (SP t-1 ) Dickey and Fuller (1979) actually consider three different regression equations that can be used to test for the presence of a unit root (Enders, 2010): SP t = (ɑ-1) SP t-1 + ε t... (1) SP t = α + (ɑ-1) SP t-1 + ε t... (2) SP t = α + (ɑ-1) SP t-1 + ρt + ε t... (3) 233

I. Awad et al. From null hypothesis the time series has no clear overall trend direction, the relevant parameters restrictions for a stochastic trend are that α = 0 and ρ = 0. The alternative of a stationary process corresponds to -2 < ρ < 0 and in this case ρ is included to model the possibility of non-zero mean of the process (Hei et al. 2004). So, the hypotheses regarding the time series data are: H O : Time series aren t stationary, α = ρ = 0 H 1 : Time series are stationary, -2 < ρ < 0 That is, when the time series data are stationary, then the data have constant means and variances, so ρ is more than -2 and less than 0, while non-stationary data, on the other hand, have the means and variances that change over time, so α = ρ = 0 and therefore there is a trend, cycles and random walks. Accordingly, Enders (2010) introduced one example of non-stationary data which are random walk. The random walk model suggests that day to day changes in the price of a stock should have a mean value of zero and the changes in stock prices should be normally distributed. Granger Causality Test Causality test of Granger is used in order to investigate whether or not there is a causal relationship between two variables in the long-run and which one causes this relationship. Thus, the Granger model is used to investigate how much of AQI can be explained by a past value of macroeconomic variables (MEV). However, to determine the relationship, Granger causality test can provide indication of the relationship between those variables. This leads to acceptance or reection of H 0 using F-test and probability. The following equation can estimate the relationship between any variables: MEV t-1 = α 1 MEV t-1 + β 1 AQI t-1 + ɛ t Granger Causality test, in this respect, is used to investigate whether or not there is a causal relationship between Al-Quds Index and macroeconomic variables and which one causes the other. Granger causality based on stationarity of time series data. That is, investigating causality involves determining whether Al- Quds Index is a function of macroeconomic variables. Marinazzo et al., (2011) stated that Granger causality analysis (Granger, 1969; Wiener, 1956) is an approach that measures the causal association and effective connectivity and can provide information about the dynamics and directionality on both variables. Testing for temporal causality between the three markets is centered on a VAR (vector autoregressive) model comprising two stationary series, x and y. This model is adopted in order to capture short run causality between the two markets. Khan et al., (2010) state that in VAR modeling the value of variable is expressed as a linear function of the past or lagged values of that variables and all other variables included in the model. Thus all variables are regarded as endogenous. The model can be written as: x t p 1 x y i t i i 1 1 q t x, t y t p 2 y x q i t i i 1 1 Where x and y are stationary variables, p and q are the lag lengths for x and y respectively and ɛ s are the stochastic error terms or shocks in the language of VAR. Testing causality between the three markets is one of the study obectives. Generally, testing causality involves using F-tests to test whether lagged information on a variable Y provides any statistically significant information about variable X in the presence of lagged X. If not, Y doesn t Granger cause X. In other words, a variable Y is said not to Granger -cause a variable X if the distribution of X, conditional on past values of X alone, equals the distribution of X, conditional on past realizations of both X and Y. If this equality doesn t hold, Y is said to Granger causes X. If Y can predict future X, over and above what lags of X itself can, then Y Granger causes X. t y, t 234

Q1 2002 Q3 2002 Q1 2003 Q3 2003 Q1 2004 Q3 2004 Q1 2005 Q3 2005 Q1 2006 Q3 2006 Q1 2007 Q3 2007 Q12008 Q3 2008 Q12009 Q3 2009 Q12010 Q3 2010 Q12011 Q3 2011 Q1 2012 Q3 2012 Q1 2002 Q3 2002 Q1 2003 Q3 2003 Q1 2004 Q3 2004 Q1 2005 Q3 2005 Q1 2006 Q3 2006 Q1 2007 Q3 2007 Q12008 Q3 2008 Q12009 Q3 2009 Q12010 Q3 2010 Q12011 Q3 2011 Q1 2012 Q3 2012 International Journal of Financial Economics In other words, if lagged information on macroeconomic variables provides any statistically significant information about Al-Quds Index in the presence of lagged observations of Al-Quds Index, then Macroeconomic variables Granger cause Al-Quds Index. Equations above are valid for testing the causality of lagged volume changes on price changes (where x is the stationary price series) and lagged price changes on volume changes (where y is the stationary volume series). As such, by using E-views we can calculate Granger causality parameters. 4. Empirical Results and Analysis Chart Analysis Graphs were used to clarify movements of the variables values along the stipulated period. Figures 2, 3, and 4 plot the quarterly values of the AQI, and macroeconomic variables- GDP, CPI, UNEMP, EXR, and IR along the stipulated time period. The time series observations of each variable have been connected together in a single line to view changes in its values and to detect trend during the stipulated period. These figures display ups and downs in each variable to indicate that the mean isn t constant and therefore their times series data isn t constant. However theses figures are likely to shed light on trends of AQI and the used macroeconomic variables that may have impact on AQI. Generally, we notice that the macroeconomic variables and AQI didn t move in the same direction. 2000 1500 1000 500 0 AQI GDP Figure 2: Quarterly values of AQI and GDP from 2002 to 2012. 160 140 120 100 80 60 40 20 0 CPI UNEMP Figure 3: Quarterly values of CPI, UNEMP from 2002 to 2012. 235

Q1 2002 Q3 2002 Q1 2003 Q3 2003 Q1 2004 Q3 2004 Q1 2005 Q3 2005 Q1 2006 Q3 2006 Q1 2007 Q3 2007 Q12008 Q3 2008 Q12009 Q3 2009 Q12010 Q3 2010 Q12011 Q3 2011 Q1 2012 Q3 2012 I. Awad et al. 12 10 8 6 4 2 0 EXR IR Regression Analysis Figure 4: Quarterly values of EXR and IR from 2002 to 2012. Regression techniques were used to describe the relationship between AQI and macroeconomic variables. According to regression output presented in the table 1 below, the overall significance indicates that macroeconomic variables explain and describe only 29.8% of AQI. The F-value is 3.223 at P-value of 0.016, which indicates that the model is significant since 0.016<0.05. However, the regression analysis indicates a weak explanatory power of the regression model since R-square has a value of about 29.8%. Table 1: Model Summary R Square Std.Error of the Estimate F Test Sig. Sum of square- Regression Sum of squares- Residual Predictors: (Constant), IR, UNEMP, EXR, GDP, CPI Dependent Variable: AQI 29.8% 192.3 3.223 0.016 595842.5 1405147.2 Table 2 shows coefficients of the variables, t-values and P-values on order to indicate the significance of the relationship for each independent variable, so the model that can describe the relationship between AQI and macroeconomic variables is: AQI = 53. 5 0. 33 GDP + 9 CPI 28. 7 UNEMP + 167 EXR 24 IR The Model presents the relationship between AQI and the macroeconomic. As shown, the AQI generally has a positive correlation with CPI, and EXR but a negative correlation with GDP, UNEMP, and IR. Variable Table 2: Regression output Unstandardized Standardized Coefficients Coefficients B Std. Error Beta T Sig. (Constant) -53.5 1292-0.041 0.967 GDP -0.33.417-0.398-0.779 0.441 CPI 9 9.85 0.617 0.917 0.365 UNEMP -28.7 11.69-0.449-2.460 0.019 EXR 167 169.75 0.327 0.984 0.331 IR -24 20.28-0.249-1.184 0.244 Dependent Variable: AQI 236

International Journal of Financial Economics T-values show that coefficient for UNEMP is significant since P-value < 0.05 but the coefficients of GDP, CPI, EXR, IR aren t significant as P-value > 0.05 and these results are in contradiction of the results of many empirical studies by Abu-Libdeh and Harasheh (2011), Hsing (2011), Hsing et al. (2012), Sohail and Hussain (2011), Binti Mohamed et al. (2011), Gnzalo and Taamouti (2011) and D.Gay, Jr, (2008). Therefore, the model can t be considered significant and doesn t explain high level of variation. Accordingly, at the 0.05 significance level we can accept H 0 that states that the Macroeconomic variables aren t related to AQI in favor of the alternative hypothesis. As a result, the macroeconomic variables haven t been considered in investment decisions. Furthermore, investors as well as policy makers should consider other variables that may explain movements of AQI along the stipulated period. Normality Test According Jarque-Bera test presented in table 3, there is a considerable evidence that the data observations of AQI, CPI, GDP, EXR aren t normally distributed since sig.>0.5 but IR and UNEMP are normally distributed at sig.<0.05. The level of normality leads to using parametric and nonparametric tests. Since most varaibles aren t normally distributed, non-parametric tests namely ADF will be used to test for stationarity as well as Granger causality for drawing the conclusion about the impact of macroeconomic variables on the AQI. Table 3: Descriptive Statistics of AQI and Macroeconomic Variables Variable/Stat Mean St. Dev. Skewness Kurtosis Jarque-Bera Prob. AQI 468.6 215.7 0.4390 3.6970 2.3066 0.315 GDP 1222.627 264.1609 0.4194 2.2787 2.2440 0.325 CPI 114.12 14.7 0.1020 1.59 3.7205 0.155 UNEMP 24.636 3.356 1.1168 4.846 15.395 0.0004 EXR 4.1345 0.422388-0.0173 1.7278 2.9693 0.226 Stationarity Test IR 3.9515 2.2404 0.8998 0.6391 6.6869 0.0353 Since this study depend on time series analysis, stationarity test is important to make data more smoothly for analysis. Table 4 shows the summary of unit root test for the data of AQI, CPI, EXR, GDP, IR, and UNEMP. As a result, Augmented Dickey Fuller test (ADF) test indicates that mean and variance seem to be constant in the time series of the variables at P- values < 0.01 in first differences so the data are stationary. Therefore, we accept the unit root alternative hypothesis of stationarity for time series data of the six variables at significance level of 1%. Table 4: Unit Root Test for Quarterly Values for AQI and Macroeconomic Variables Augmented Dickey Fuller test Variable At Level First Difference T-Stat. P-value T-Stat. P-value AQI -2.3776 0.0224-4.678317 0.000 GDP -0.285335 0.7769-6.521083 0.000 CPI 0.214062 0.8316-4.987 0.000 UNEMP -3.216894 0.0026-5.993494 0.000 EXR -2.105817 0.0417-4.675166 0.000 IR -2.633569 0.0121-5.747654 0.000 237

I. Awad et al. In this step we use the Granger Causality test to check the causal relationship between time series. The Pairwise Granger Causality test uses the F- test. We use this test as an indicator for the direction of the relationship. Granger Causality Test Stationarity test is an initial step to conduct Granger causality relationship. Accordingly, ADF test suggest that all time series of the variables are stationary at first difference at 1% level of significance, so we have no restrictions in conducting Granger causality tests between the AQI and macroeconomic variables. Correlation matrix has been used to indicate whether the variables are correlated with each other. The correlation coefficients for AQI and macroeconomic variables are less than 1, which indicates that the variables seem to be not correlated with other variables and amongst each other. Further, the variables moved independently and not correlated with the others. Therefore, there are no linkages between AQI and macroeconomic variables. Table 5: Correlation Matrix for the Three Stock Market Indices AQI CPI EXR GDP IR UNEMP AQI 1.00000 0.35400-0.29937 0.35820-0.42043-0.48078 CPI 0.35400 1.00000-0.87993 0.94472-0.74874-0.47893 EXR -0.29937-0.87993 1.00000-0.77069 0.68726 0.48964 GDP 0.35820 0.94472-0.77069 1.00000-0.70690-0.55719 IR -0.42043-0.74874 0.68726-0.70690 1.00000 0.48129 UNEMP -0.48078-0.47893 0.48964-0.55719 0.48129 1.00000 The results of correlation matrix need to be further verified for the direction of influence by the Granger causality test. Further, ADF test indicates that time series data of the variables are stationary at first difference at 1% level of significance, so cointegration isn t applicable in this case. Therefore, Granger causality test has been conducted to capture the degree and the direction of long term causality relationship between the variables. Table 6 presents the results of Granger Causality test for this purpose. Table 6: Results of Pairwise Granger Causality Test Pairwise Granger Causality Tests Variable F-Statistic Probability CPI AQI 0.00774 0.99229 AQI CPI 0.11221 0.89416 EXR AQI 0.10733 0.89851 AQI EXR 0.04094 0.95993 GDP AQI 0.06406 0.93806 AQI GDP 1.31212 0.28148 IR AQI 1.18391 0.31741 AQI IR 0.17313 0.84171 UNEMP AQI 1.45570 0.24628 AQI UNEMP 0.31823 0.72940 Significance levels of 1%, 5% and 10% On the basis of Granger Causality output, AQI does not Granger causes the macroeconomic variables that been tested, and the macroeconomic variables also do not Granger cause the AQI since P-value > 0.1 for all cases. 238

International Journal of Financial Economics So the null hypotheses are accepted under the two cases. Thus, there is no mutual causal relationship between macroeconomic variables and AQI. Accordingly, there is no feedback relationship among these variables. 5. Conclusions and Policy Implications This study examined causality relationship between Al-Quds Index and the macroeconomic variables to investigate whether or not the AQI is explained by those variables. The study is likely to provide policy of recommendations to both policymakers and researchers not only in Palestine but also in the region. According to regression analysis, the study found that the AQI isn t significantly linked with macroeconomic indicators, thus the macroeconomic variables cannot explain the behavior of the AQI, but not in the same direction indicated in the researches reviewed due to the inefficiency of the Al Quds index. Added to this, Jarque-Bera test indicates that most variables are not normally distributed since P-value > 0.05 level of significance, so that ADF test has been used to test data stationarity. ADF test indicates that time series data of the variables are stationary at first difference at 1% level of significance, so cointegration isn t applicable in this case. Further, the correlation coefficients for AQI and macroeconomic variables are less than 1, which indicates that the variables seem to be not correlated with other variables and amongst each other. Once ADF and correlation test have been conducted, Granger Causality test has been conducted to capture the degree and the direction of long term causality relationship between the variables. Granger Causality output indicates that AQI does not Granger causes the macroeconomic variables that been tested, and the macroeconomic variables also do not Granger cause the AQI since P-value > 0.1 for all cases. Therefore, there is no feedback causal relationship among AQI and macroeconomic variables. The reason behind these results might be lack of information, or most probably lack of liquidity and depth in the Palestinian market as well as political and economic circumstances. It is important to keep in mind that the Palestinian stock market is still young and at its early stages and is inefficient; especially there isn t enough liquidity in this market in order to allow it to respond to macroeconomic forces. This study is important for investors, researchers and policy makers. The Palestinian investor is advised to follow news about the stock market and the economy as a whole in order to minimize risk and avoid losses. Further, detailed statistics should be collected and provided to enable researchers to conduct in-depth studies on the impact and role of PSE that will contribute to formulating appropriate policies and take the right steps that lead to attracting foreign investments. Decision makers should pay more attention about the relationship between macroeconomic variables and stock prices in PEX when setting rules and regulations that affect any one of them. It also should promote Palestinian shares in external markets, particularly in countries such as the Gulf States that host large Palestinian communities and continue to issue the Annual Directory of listed public shareholding companies, and distribute it to investors in other counties through Palestinian Embassies and Representative Offices. Lastly, other variables that really affect AQI should be considered for future research. References Ahmet Büyükşalvarcı (2010). The Effects of Macroeconomics Variables on Stock Returns: Evidence from Turkey, European Journal of Social Sciences, Volume 14, Number 3 (2010). Haneen Abu-Libdeh, Murad Harasheh (2011). Testing for correlation and causality relationships between stock prices and macroeconomic variables The case of Palestine Securities Exchange. International Review of Business Research Papers, Vol. 7. No. 5. September 2011. Pp. 141-154. 239

I. Awad et al. Jes us Gonzalo, Abderrahim Taamouti (2011). The Reaction of Stock Market Returns to Anticipated Unemployment, Financial support from the Spanish Ministry of Education through grants SEJ 2007-63098 is acknowledged. Mahmood, W & Dinniah, N (2009). Stock Returns and Macroeconomics Variables Evidence from the Six Asian-Pacific Countries, International Research Journal of Finance and Economics, vol. 30, Pp 154-162. McQueen, Grant and Vance V. Roley (1993). Stock Prices, News and Business Conditions, Review of Financial Studies, 92, 307-328. Mofleh Ali Mofleh Alshogeathri(2011). Macroeconomic Determinants Of The Stock Market Movements : Empirical Evidence From The Saudi Stock Market, PHD Thesis Department of Economics, Kansas State University. Nadeem Sohail, Zakir Hussain (2011). The Macroeconomic Variables and Stock Returns in Pakistan: The Case of KSE 100 Index, International Research Journal of Finance and Economics, ISSN 1450-2887 Issue 80 (2011),Euro Journals Publishing, Inc. 2011. Pilinkus, D (2009). Stock Markets and macroeconomic variables: evidences from Lithuania, Journal Of Economics and Management, vol.14, pp 884-893. Robert D.Gay,Jr, (2008). Effect Of Macroeconomic Variables On Stock Market Returns For Four Emerging Economies Brazil, Russia, India, And China. International Business & Economics Research Journal, Volume 7, Number 3. Suhana binti Mohamed Alizah binti Ali, Nuriza binti Abdullah, and Katrina binti Azman (2011). Malaysian Sectoral Indices VS Macroeconomic Factors, Any Correlation?: Evidence From Malaysia, International Conference on Economics and Finance Research, IPEDR vol.4 (2011) (2011)IACSIT Press, Singapore. Yu Hsing (2011). Effects of Macroeconomic Variables on the Stock Market: The Case of the Czech Republic. Theoretical and Applied Economics, Volume XVIII (2011), No. 7(560), pp. 53-64. Yu Hsing (2011). The Stock Market and Macroeconomic Variables in a BRICS Country and Policy Implications. International Journal of Economics and Financial Issues, Vol. 1, No. 1, 2011, pp. 12-18, www.econournals.com. Yu Hsing (2011). Macroeconomic Determinants of the Stock Market Index and Policy Implications: The Case of a Central European Country. Eurasian Journal of Business and Economics, 2011, 4 (7), 1-11. Yu Hsing, Michael C.Budden, and Antoinette S.Phillips(2012). Macroeconomic Determinants of the Stock Market Index for a Maor Latin American Country and Policy Implications. Business and Economic Research, ISSN 2162-4860, 2012, Vol.2, No.1. Zvi Bodie, Alex Kane, and Alan J. Marcus, (2009). Investments, Eighth Edition, International Edition 2009. Chapter 1, Pp 5. Website of CBS: http://www.cbs.gov.il. Website of PCBS: http://www.pcbs.gov.ps. Website of PEX: http://www.pex.ps/psewebsite/english/default.aspx 240

International Journal of Financial Economics 241