acroeconomic variables play an important role in the performance of stock market returns. Numerous studies document that there are link between macroeconomic variables and equity returns. It is found that changes in the macroeconomic environment affect the price of share. According to the arbitrage pricing theory the relation between stock returns and certain macroeconomic variables has been established (Ross-1976). In addition, some studies concerning multifactor models frequently incorporate certain macroeconomic variables as explanatory factor of the expected returns (Bilson et. al. 2001). A potential investor and portfolio manager looks at such a stock market where macroeconomic variable are moves sense of direction. It is very interesting to invest stock market but a very risky trench of investment. So, potential investors always try to predict the trends of stock market prices to obtain maximum benefits and minimize the E-mail : [email protected] future risks. Being concerned with the relationship between stock market returns and macroeconomic variables, investors might guess how stock market behaved if macroeconomic indicators such as exchange rate, industrial productions, interest rate, consumer price index and money supply fluctuate (Hussainey and Ngoc, 2009). Macroeconomic indicators are compositions of data which frequently used by the policy makers and investors for gathering knowledge of current and upcoming investment priority. The present studies have concentrated on two developed countries' stock markets such as Germany and the United Kingdom and will try to find out the relationship between stock market returns and certain macroeconomic variables in Frankfurt stock exchange and the London stock exchange.

The rest of the study is structured as follows: section two highlights on related literature, section three concentrates on methodology and description of the dataset, section four discusses the empirical results and finally, section five draws a conclusion to the study.

In globalized economy there are various ways financial market especially the stock market and the macro-economy have been related in the literature. In recent past, longstanding academic studies evidence that macroeconomic indicator affects stock prices. We find plenty of research on how the macroeconomic indicators affect the stock market. In 1981, Fama established a relationship among stock prices and macroeconomic indicators. He found that expected nominal inflation is negatively correlated in real activity and the reality is that the changing inflation has positive relation to returns on the stock market. Later studies support the Fama's (1981) hypothesis. Geske and Roll (1983) emphasized on the importance of policy responses in explaining stock returns. In 1987 Kaul also emphasized the same.

Errunza and Hogan (1998) examined whether the variability of a set of monetary and real macroeconomic factors can explain the variation of the some European stock market volatility. Employing a Vector-auto Regression (VAR), they found evidence to support that monetary instability is a significant factor for Filis (2010) found that there is no causal relationship between Greek stock market and industrial production during the period spanning from January 1996 to June 2008 using multivariate VAR model. He also argued, stock market and oil prices exercise a positive impact on Greek consumer price index in the long-run. Daly and Fayyad (2011) examined, the relationship between Gulf Cooperation Council (GCC) countries, the UK and the US stock market returns and oil price by employing DCV and VAR analysis during the period September 2005 to February 2010 and find that when oil prices increase sharply it predicts the USA, UAE and Kuwait but not the UK, Oman, Bahrain and Qatar.

There are little segmentation observed between emerging and developed market stock returns. The volatility of developed economies' stock returns is less than the volatility of emerging market stock returns. The volatility of emerging market is changed by local macroeconomic variables as well as international macroeconomic variables. Abugri (2008) The correlations between stock market returns and the macroeconomic variables are different. A positive correlation is evident between the DAX30 and the macro-economic variables with the exception of bond;the correlation (table-1) between the UK price index and the macroeconomic variables are fairly strong with the exception of CPI and MS. In research, the data sources, data description and the methodology need to be specified. The methodology needs to be cautiously designed to obtain realistic results. The methodological design employed in this study consists of unit root tests; Johansen cointegration test, VECM based Granger causality, variance decomposition analysis and impulse response analysis.

The empirical investigation has been carried out in the case of the United Kingdom and German stock market returns and selective macroeconomic variables. The data used under the study are monthly data from February 1999 to January 2011. The UK and German stock prices is the end-of-period closing share price indices.

The stock indices are DAX30 of Frankfurt stock exchange and FTSE100 of London stock exchange. These stock price indices and the chosen macroeconomic variables such as broad money supply (MS), exchange rates, treasury bill rates (Representing interest rate for UK), bond rate (Representing interest rate for Germany) are obtained from the Data Stream.

Consumer price index (CPI) representing the rate of inflation and Industrial Production Index (IP) representing the economic activity are sourced from OECD data bank.

The stock market returns of Germany and the UK are shown a high level of time varying correlation. If we have a close look towards German and the UK stock markets return (figure-4.1), we observe that these two developed economies stock market returns are closely correlated in the sample period except late 2000. The first step of the methodological process involves a test for stationarity as the variables to be used in this paper are time series which are usually nonstationary. We employed Augmented Dickey-Fuller(ADF) and Phillips-Perron (PP) tests for unit root. If the variables are stationary in level, they are said to be integrated of order 0 that is I(0). On the other hand, if the said variables become stationary after first differencing are said to beI(1). c) Johansen Multivariate Co-integration Test:

Co-ingration refers to the situation where the nonstationary time series of the same order exist a longrun relationship. After determing the order of integration of each variables, we perform Johansen co-integration tests whether there is a cointegrating relationship between stock returns and chosen five macroeconomic variables in Germany and the UK. The mathematical form of Johansen cointegration test is given below: Where = k vector of endogenous variables, a vector of deterministic variables, = a vector of innovations. The model (i) may be re-written as a vector auto regression (VAR) following way In equation (ii) the vector and are I(1) variables. Therefore, the long run relationship among will be determined by the rank of , if r= 0 the n the equation (ii) reduce to a VAR model of p-th order and in this case the macroeconomic variables in level do not have any co-integrating vector. On the other hand, If the rank 0<r<n then there is a possibility of existing n×r matrices namely ? and ? and it can be written such that The Johansen co-integration test estimate the ma trix from a n unre s trite d VAR a nd a ls o te s t whe the r we can reject the restrictions implied by the reduced rank of using either the trace statistic or the maximum eigen value statistic (Wickremasinghe, 2011). The trace statistic and the maximum eigenvalue statistic is determined using the following equations

Where T= Number of observations, = Estimated values of characteristic roots ranked from largest to smallest and r= 0,1,2,......n-1. It is well known that the co-integration test is Lag sensitive. This study follows the Akaike Information criterion (AIC) and Schwarz Bayesian Criterion (SBC) to select the number of appropriate lags.

If thre exists a co-integration relationship between the stock returns and macroeconomic variables then there is a possibility of causality among the variables at least one direction (Engle and Geanger, 1987). If we consider (stock market indices) and (macroeconomic variables) as two different time series then the error correction model express as following way:

Where is the difference operator, n and m are the lag lengths of the variables, is the re s idua l from the co-integrating equation.

a nd a re the disturbance terms. From equation (vi) and (vii) we can examine the statistical significance of the error correction term by separate t-test and the joint significance of the lags of each explanatory variables by -test.

The standard Granger causality analysis interpreted within the sample period only. In this regard, variance decomposition analysis could be an important tool to make proper inference regarding the causal relationships beyond the sample period. Actually, Variance Decomposition indicates the percentage of the forecast error variance in one variable that is due to errors in forecasting itself and each of the other variables (Tarik, 2001).

The impulse response function is designed to infer how each variable responds at different time horizon to an earlier shock in that particular variable and to shocks in other macroeconomic variables. Particularly, we investigate the response of the DAX30/ FTSE100 to one standard deviation shocks to the equation for DAX30/FTSE100 and macroeconomic variables and also the response of macroeconomic variables to one standard deviation to the equation for the DAX30/FTSE100.

IV.

The unit-root test is performed on the UK and German time series to determine whether the time series is stationary. We employed both the ADF and PP unit root tests. The findings of the unit-root test are shown in Table 2. The results indicate that all the variables show unit roots at natural log level and stationary at its first differences. Therefore, the variables are integrated of order one that is I(1). Thus, we are able to investigate the long-run equilibrium relationship among the macroeconomic variables. The Johansen co-integration test results particularly trace statistic and eigenvalue statistic are presented in table-3.

The result represents that both DAX30 and FTSE100 are co-integrated with corresponding macroeconomic variables. Thus, the results implies that there is long run equilibrium relationship between the stock market prices and the five macroeconomic variables in Germany and the UK during the periods under the present study. and Business Research Volume XII Issue XVI Version I

As we found cointegrating relationship for both the countries,we proceed to investigate the error correction models. The results obtain from Error Correction Mechanism (ECM) specification represented by model (vi) and (vii) is depicts in table-4. According to the results we can see the four sorts of causal relationship such as short-run, long-run, no causality and both short and long run causal relationship. We find there are three short-run, two long-run and one short and long run casual relationships for Germany. The short run causality run from DAX30 to CPI, from money supply (MS) to DAX30 and from industrial production (IP) to DAX30. The long-run causality runs from CPI to DAX30 and from exchange rates to DAX30.

There is only one short and long-run relationship, that from the DAX30 to industrial production. For the United Kingdom , We find there are five short-run, one long-run and two short and long run casual relationships. The short run causality runs from FTSE100 to Tbill, from FTSE100 to MS, from FTSE100 to exchange rate, exchange rate to FTSE100 and FTSE100 to industrial production. The long-run causality runs from CPI to FTSE100 . The short and long-run causal relationship runs from FTSE100 to CPI, from MS to FTSE100 and from IP to FTSE100 The results of variance decomposition analysis of Germany are presented in table-5 and 6. The table-5 decomposes with the stock market indices of Germany and the macroeconomic variables. The variance decomposition analysis was employed to supplement the Granger causality results to reinvestigate the out of sample impact. The results provided in columns 2-6 of table-5 indicates how much of the DAX30's own shock is explained by movements in its own variance and the chosen macroeconomic variables over the 60 months forecast horizon. According to the results, shown in table-5, the amount of variance of the DAX30 explained by own goes down when the time horizon increased up to 60 months. At horizon one, all variance in the DAX30 is explained by own. At horizon 60, 85% of DAX30 variance is explained by itself. This indicates that at longer horizons, the variance of DAX30 may be caused by variance of other macroeconomic variables especially by money supply and industrial production. At horizon 24, the IP explains 5.93% of the variances of the DAX30. When the time horizon goes up, the actual The result presented in table-4 indicate that, there is a unidirectional causality running from DAX30 to CPI and MS to DAX30, IP to DAX30, CPI to DAX30, exchange rate to DAX30. Based on the above result, we can conclude that the share price of Germany (DAX30) can be predicted from certain macroeconomic variables. Thus, the German stock market index does behave according to the predictions of the efficient market hypothesis (Wickremasinghe, 2011). goes down when the time horizon increased up to 60 months. At horizon one all variance in the FTSE100 is explained by own. At horizon 60, 84% of FTSE100 variance is explained by itself. This indicates that at longer horizons, the variance of FTSE100 may be caused by variance of other macroeconomic variables especially by exchange rate and industrial production. At horizon 24, the IP explains 6.35% of the variances of the FTSE100. When the time horizon goes up, the actual amount of variance of the FTSE100 explained by the IP also goes up. The other variable may cause in the FTSE100 is exchange rate. At horizon 48, 5.48% of the variance in the FTSE100 is explained by exchange rate. The consumer price index (CPI), bond and money supply play little role in explaining the variance of the FTSE100.

© 2012 Global Journals Inc. (US)

and Business Research Volume XII Issue XVI Version I Table 7 : Variance Decomposition Analysis Results for FTSE100.

The percentage of forecast variance in macroeconomic variables explained by the innovations of FTSE100 is presented in table-8. Table-8 indicates that the FTSE100 explains very little forecast variance of the money supply (MS) and CPI. The percentage of forecast variance in MS by FTSE100 is 3.71% in horizon 12, however when the time horizon increase then percentage of forecast variance in MS by FTSE100 is goes down. The macroeconomic variable whose variance is explained significantly by the FTSE100 is Tbill, IP and exchange rate. For example, the FTSE100 explains 24.85%, 18.15% and 11.24% of the variance in the T-bill, IP and exchange rate respectively at the forecast horizon 6. The result presented in table-4 indicate that, there is a unidirectional causality running from FTSE100 to T-bill, FTSE100 to MS and CPI to FTSE100, MS to FTSE100, IP to FTSE100. Based on the above result, we can conclude that the share price of the UK (FTSE100) can be predicted from certain macroeconomic variables. Thus, the UK stock market index does behave according to the predictions of the efficient market hypothesis (Wickremasinghe, 2011). Figure-3 indicates impulse response of FTSE100 to one standard deviation shock in the equations for FTSE100 and five macroeconomic variables and also the impulse response of five macroeconomic variables to one standard deviation shock in the equation for FTSE 100. A standard deviation shock in the equation for the FTSE 100 increases the FTSE 100 until horizon five, after which a standard deviation shock to the equation for FTSE100 does not produce any volatility in the FTSE100. Response of FTSE100 to MS has negative impact.

Response of FTSE100 to IP, CPI to FTSE100 and IP to FTSE100, Tbill to FTSE 100, Exrate to FTSE 100 has positive impact. The response of MS to FTSE 100 shows volatiltility up to 18th horizon, after which there is no volatility observed. V.

This study examined the causal relationship between stock prices and a set of selected macroeconomic variables in Germany and the United Kingdom. We investigated both short and long-term relationship between stock prices and the chosen macroeconomic determinants. We employed both the ADF and PP unit root tests. We carefully selected the deterministic components in the Johansen cointegration test. The results of the Johansen cointegration test indicate that there is co-integrating relationship between the stock prices and macroeconomic determinants in the case of German and the UK markets. After establishing cointegration The result of the study are consistent with the majority of the relevant literature, implies the existence of short run interactions and long term causal relationship between both Germany and the UK stock markets and the respective fundamentals. We find there are three short-run, two long-run and one short and long run casual relationships for Germany. The short run causality runs from DAX30 to CPI, from money supply (MS) to DAX30 and from industrial production (IP) to DAX30. The lon-run causality runs from CPI to DAX30 and from exchange rate to DAX30. There is only one short and long-run relationship, that is from the DAX30 to industrial production. For the United Kingdom , We find that there are five short-run, one long-run and two short and long run casual relationships. The short run causality run from FTSE100 to Tbill, from FTSE100 to MS, from FTSE100 to exchange rate, exchange rate to FTSE100 and FTSE100 to industrial production. The lonrun causality runs from CPI to FTSE100 . The short and long-run causal relationship runs from FTSE100 to CPI, from MS to FTSE100 and from IP to FTSE100. These results indicate that stock prices in Germany and the UK can be predicted using certain macroeconomic varibles.

The analysis of variance decomposition for Germany found that, at short term horizons most of the forecast horizons of the stock prices are explained by the stock price itself. However, in the long run horizons MS and IP play an important role in explaining the forecast variance in stock prices. When macroeconomic determinants are concerned, the stock prices are able to explain the forecast variance of the IP, Bond and CPI. Furthermore, The analysis of variance decomposition for the United Kingdom market found that, at short term horizons most of the forecast horizons of the stock prices are explained by the stock price itself. However, in the long run horizons Exchange rate and IP play significant roles in explaining the forecast variance in stock prices. When macroeconomic determinants are concerned, the stock prices are able to explain the forecast variance of the IP and T-bill.

The impulse response function of the DAX30 to a standard deviation shock given to the equation for five macroeconomic determinants found that a shock to the macroeconomic variable equations responses from the DAX30 only at the shorter horizons. We also examined whether a stock given to the DAX30 generated any response from macroeconomic determinants. We found that, a standard deviation shock in the equation for the DAX30 increases the DAX30 until horizon six, after which a standard deviation shock to the equation for DAX30 does not produce any volatility in the DAX30. Response of DAX30 to CPI, DAX30 to MS and exchange rate to DAX30 has negative impact. Responses of DAX30 to IP, DAX30 to Bond, CPI to DAX30 and IP to DAX30 has positive impact. Furthermore, The impulse response function of the FTSE100 to a standard deviation shock given to the equation for five macroeconomic determinants found that a shock to the macroeconomic variable equations responses from the FTSE100 only at the shorter horizons. We also examined whether a stock given to the FTSE100 generated any response from macroeconomic determinants. We found that, a standard deviation shock in the equation for the FTSE 100 increases the FTSE 100 until horizon five, after which a standard deviation shock to the equation for FTSE100 does not produce any volatility in the FTSE100. Responses of FTSE100 to MS has negative impact. Responses of FTSE100 to IP, CPI to FTSE100 and IP to FTSE100, Tbill to FTSE 100, Exrate to FTSE 100 has positive impact. The response of MS to FTSE 100 shows volatiltility up to 18th horizon, after which there is no volatility observed. The findings of co-integration, shortrun and long-run causal relationship between stock indices and certain macroeconomic variables in our research help policy makers, investors and portfolio manager in efficient investment decision making in both the German and the UK stock markets.