International Research Journal of Finance and Economics ISSN 1450-2887 Issue 14 (2008) EuroJournals Publishing, Inc. 2008 http://www.eurojournals.com/finance.

htm

Aggregate Economic Variables and Stock Markets in India

Shahid Ahmed Economist, India Programme, UNCTAD, New Delhi (India) Department of Economics, Jamia Millia Islamia (Central University), New Delhi E-mail: shah_ec_jmi@yahoo.co.in Abstract This study investigates the nature of the causal relationships between stock prices and the key macro economic variables representing real and financial sector of the Indian economy for the period March, 1995 to March, 2007 using quarterly data. These variables are the index of industrial production, exports, foreign direct investment, money supply, exchange rate, interest rate, NSE Nifty and BSE Sensex in India. Johansen`s approach of cointegration and Toda and Yamamoto Granger causality test have been applied to explore the long-run relationships while BVAR modeling for variance decomposition and impulse response functions has been applied to examine short run relationships. The results of the study reveal differential causal links between aggregate macro economic variables and stock indices in the long run. However, the revealed causal pattern is similar in both markets in the short run. The study indicates that stock prices in India lead economic activity except movement in interest rate. Interest rate seems to lead the stock prices. The study indicates that Indian stock market seems to be driven not only by actual performance but also by expected potential performances. The study reveals that the movement of stock prices is not only the outcome of behaviour of key macro economic variables but it is also one of the causes of movement in other macro dimension in the economy. Keywords: Cointegration Test, T-Y Granger Causality Test, Variance Decomposition, Impulse Response Function JEL Classification Codes: G1, F4, C22.

Introduction
The movement of stock indices is highly sensitive to the changes in fundamentals of the economy and to the changes in expectations about future prospects. Expectations are influenced by the micro and macro fundamentals which may be formed either rationally or adaptively on economic fundamentals, as well as by many subjective factors which are unpredictable and also non quantifiable. It is assumed that domestic economic fundamentals play determining role in the performance of stock market. However, in the globally integrated economy, domestic economic variables are also subject to change due to the policies adopted and expected to be adopted by other countries or some global events. The common external factors influencing the stock return would be stock prices in global economy, the interest rate and the exchange rate. For instance, capital inflows and outflows are not determined by domestic interest rate only but also by changes in the interest rate by major economies in the world. Recently, it is observed that contagion from the US sub prime crisis has played significant movement

142

International Research Journal of Finance and Economics - Issue 14 (2008)

in the capital markets across the world as foreign hedge funds unwind their positions in various markets. Other burning example in India is the appreciation of currency due to higher inflow of foreign exchange. Rupee appreciation has declined stock prices of major export oriented companies. Information technology and textile sector are the example of falling stock prices due to rupee appreciation. The modern financial theory focuses upon systematic factors as sources of risk and contemplates that the long run return on an individual asset must reflect the changes in such systematic factors. This implies that securities market must have a significant relationship with real and financial sectors of the economy. This relationship generally viewed in two ways. The first relationship views the stock market as the leading indicator of the economic activity in the country, whereas the second focuses on the possible impact the stock market may have on aggregate demand, particularly through aggregate consumption and investment. The former case implies that stock market leads economic activity, whereas the latter suggests that it lags economic activity. Knowledge of the sensitivity of stock market to macro economic behaviour of key variables and vice-versa is important in many areas of investments and finance. This research may be useful to understand this relationship. From the beginning of the 1990s in India, a number of measures have been taken for economic liberalization. At the same time, large number of steps has been taken to strengthen the stock market such as opening of the stock markets to international investors, regulatory power of SEBI, trading in derivatives, etc. These measures have resulted in significant improvements in the size and depth of stock markets in India and they are beginning to play their due role. Presently, the movement in stock market in India is viewed and analysed carefully by large number of global players. Understanding macro dynamics of Indian stock market may be useful for policy makers, traders and investors. Results may reveal whether the movement of stock prices is the outcome of something else or it is one of the causes of movement in other macro dimension in the economy. The study also expects to explore whether the movement of stock market are associated with real sector of the economy or financial sector or both. In this context, the objective of this paper is to explore such causal relations for India as there is a little work for this period. To the best of our knowledge no in-depth such analysis of macro dynamics is available in the literature for the period of 1995 to 2007. The paper is organised in the following sections. Section II provides review of selected literature on the causal relationship between stock prices and macro variables. Section III discusses the data and explains the methodology for testing the stationarity, the existence of cointegration, and the direction of causality. Section IV reports the results and their interpretation. Finally, Section V provides concluding remarks.

II. Review of Selected Literature

Varying evidences of causal links of stock returns and macro variables have been found in the literature using various asset pricing specifications. In the literature, widely popular CAPM has been severely challenged since returns can be predicted from other financial factors. This has led to the development and testing of various alternative asset pricing specifications, such as the Arbitrage Pricing Theory (APT) and Present Value Model (PVM). In the context of macro dynamics of stock returns, APT assumes that returns are generated by a number of macroeconomic factors. It allows multiple risk factors to explain asset returns. Chen, Roll and Ross (1986) have argued that stock returns should be affected by any factor that influences future cash flows or the discount rate of those cash flows. In an empirical investigation they found that the yield spread between long and short term government bonds, expected inflation, unexpected inflation, nominal industrial production growth and the yield spread between corporate high and low grade bonds significantly explain stock market returns. An alternative way of linking macroeconomic variables and stock prices is the discounted cash flow or present value model (PVM). This model relates the stock price to future expected cash flows and the future discount rate of the cash flows. Again, all macroeconomic factors that influence future expected cash flows or the discount rate by which the cash flows are discounted should have an

International Research Journal of Finance and Economics - Issue 14 (2008)

143

influence on the stock price. The advantage of the PVM model is that it can be used to focus on the long run relationship between the stock market and macroeconomic variables. In the literature, various theoretical reasons have been explained linking behaviour of stock prices and key macro economic variables. For instance, Friedman (1988) suggests wealth effect and substitution effect as the possible channels through which stock prices might directly effect money demands in the economy. Friedman (1988) expected that the wealth effect will dominate and thus the demand for money and stock prices to be positively related. The theoretical basis to examine the link between stock prices and the real variables are well established in economic literature, e.g., in Baumol (1965) and Bosworth (1975). The relationship between stock prices and real consumption expenditures, for instance, is based on the life cycle theory, developed by Ando and Modigliani (1963), which states that individuals base their consumption decision on their expected life time wealth. Part of their wealth may be held in the form of stocks linking stock price changes to changes in consumption expenditure. Similarly, the relationship between stock prices and investment spending is based on the q theory of James Tobin (1969), where q is the ratio of total market value of firms to the replacement cost of their existing capital stock at current prices. In retrospect of the literature, a number of hypotheses also support the existence of a causal relation between stock prices and exchange rates. For instance, goods market approaches (Dornbusch and Fischer, 1980) suggest that changes in exchange rates affect the competitiveness of a firm as fluctuations in exchange rate affects the value of the earnings and cost of its funds as many companies borrow in foreign currencies to fund their operations and hence its stock price. An alternative explanation for the relation between exchange rates and stock prices can be provided through portfolio balance approaches that stress the role of capital account transaction. Like all commodities, exchange rates are determined by market mechanism, i.e., the demand and supply condition. A blooming stock market would attract capital flows from foreign investors, which may cause an increase in the demand for a countrys currency. The reverse would happen in case of falling stock prices where the investors would try to sell their stocks to avoid further losses and would convert their money into foreign currency to move out of the country. There would be demand for foreign currency in exchange of local currency and it would lead depreciation of local currency. As a result, rising (declining) stock prices would lead to an appreciation (depreciation) in exchange rates. Moreover, foreign investment in domestic equities could increase over time due to benefits of international diversification that foreign investors would gain. Furthermore, movements in stock prices may influence exchange rates and money demand because investors wealth and liquidity demand could depend on the performance of the stock market. Economic theories suggest causal relations between stock prices and exchange rates, existing evidence also provides relatively stronger relationship between stock price and exchange rate. Ma and Kao (1990) find that a currency appreciation negatively affects the domestic stock market for an export-dominant country and positively affects the domestic stock market for an import-dominant country, which seems to be consistent with the goods market theory. Bahmani and Sohrabian (1992) found a bi-directional causality between stock prices measured by the Standard & Poor's 500 index and the effective exchange rate of the dollar, at least in the short run. The co-integration analysis revealed no long run relationship between the two variables. Similarly, Abdalla and Murinde (1996) investigate interactions between exchange rates and stock prices in the emerging financial markets of India, Korea, Pakistan and the Philippines. The results of the granger causality tests results show uni-directional causality from exchange rates to stock prices in all the sample countries, except the Philippines. Ajayi and Mougoue (1996), using daily data for eight countries, show significant interactions between foreign exchange and stock markets, while Abdalla and Murinde (1997) document that a countrys monthly exchange rates tends to lead its stock prices but not the other way around. Pan, Fok & Lui (1999) used daily market data to study the causal relationship between stock prices and exchange rates and found that the exchange rates Granger-cause stock prices with less significant causal relations from stock prices to exchange rate. They also find that the causal relationship have been stronger after the Asian crisis.

144

International Research Journal of Finance and Economics - Issue 14 (2008)

Chen, Roll, and Ross [1986], Bodie [1976], Fama [1981], Geske and Roll [1983], Pearce and Roley [1983], Pearce [1985] and many papers have tried to show empirical associations between macroeconomic variables and security returns. Bodie [1976], Fama [1981], Geske and Roll [1983], and Pearce and Roley [1983], Pearce [1985] document a negative impact of inflation and money growth on equity values. Many experts however believe that positive effects will outweigh the negative effects and stock prices will eventually rise due to growth of money supply (e.g., Mukherjee and Naka, 1995). They argue that a change in the money supply provides information on money demand, which is caused by future output expectations. If the money supply increases, it means that money demand is increasing, which, in effect, signals an increase in economic activity. Higher economic activity implies higher cash flows, which causes stock prices to rise. Bernanke and Kuttner (2005) argue that the price of a stock is a function of its monetary value and the perceived risk in holding the stock. A stock is attractive if the monetary value it bears is high. On the other hand, a stock is unattractive if the perceived risk is high. The authors argue that the money supply affects the stock market through its effect on both the monetary value and the perceived risk. Money supply affects the monetary value of a stock through its effect on the interest rate. The authors believe that tightening the money supply raises the real interest rate. An increase in the interest rate would in turn raise the discount rate, which would decrease the value of the stock as argued by the real activity theorists. The impact of real sector macro variables on equity returns has been much more difficult to establish. Mukherjee and Naka (1995) reveal that cointegration relation existed and positive relationship was found between the Japanese industrial production and stock return. However, Cutler, Poterba, and Summers [1989] (CPS) find that Industrial Production growth is significantly positively correlated with real stock returns over the period 1926-1986, but not in the 1946- 85 sub-period. In Indian context, Bhattacharya and Mukherjee (2002) studied the nature of the causal relationship between stock prices and macro aggregates for the period of 1992-93 to 2000- 2001.Their results show that there is no causal relationship between stock price and macro economic variables like money supply, national income and interest rate but there exists a two way causation between stock price and rate of inflation. Their results also indicate index of industrial production lead the stock price. As discussed above, literature reveals differential causal pattern between key macro economic variables and stock prices. This relationship varies in a number of different stock markets and time horizons in the literature. This paper will add to the existing literature by providing robust result, which is based on more than one technique, about causal links for the longer period, i.e., 12 years quarterly data.

III. Empirical Methodology and Data

Time series analysis must be based on stationary data series for drawing useful inferences. Broadly speaking a data series is said to be stationary if its mean and variance are constant (non-changing) over time and the value of covariance between two time periods depends only on the distance or lag between the two time periods and not on the actual time at which the covariance is computed. The correlation between a series and its lagged values are assumed to depend only on the length of the lag and not when the series started. This property is known as stationarity and any series obeying this is called a stationary time series. Three unit root tests have been applied to test whether a series is stationary or not. Stationarity condition has been tested using Augmented Dickey Fuller (ADF) and Phillips Perron (PP) tests. [Dickey and Fuller (1979, 1981), Gujarati (2003), Phillips and Perron (1988), Enders (1995)]. For testing null hypothesis of stationarity, KPSS test has also been applied for robustness [Kwiatkowski, Phillips, Schmidt. And Shin(1992)].

International Research Journal of Finance and Economics - Issue 14 (2008) Augmented Dickey Fuller Test

145

Augmented Dickey-Fuller (ADF) test has been carried out which is the modified version of Dickey Fuller (DF) test. ADF makes a parametric correction in the original DF test for higher-order correlation by assuming that the series follows an AR (p) process. The ADF approach controls for higher-order correlation by adding lagged difference terms of the dependent variable to the right-hand side of the regression. The Augmented Dickey-Fuller test specification used here is as follows:
p

yt

yt

1 i 1

yt

ut

(1)

Phillips-Perron (PP) Test Phillips and Perron (1988) adopts a nonparametric method for controlling higher-order serial correlation in a series. The test regression for the Phillips-Perron (PP) test is the AR (1) process. While the ADF test corrects for higher order serial correlation by adding lagged differenced terms on the right-hand side, the PP test makes a correction to the t-statistic of the coefficient from the AR(1) regression to account for the serial correlation in ut . The correction is nonparametric. The advantage of Phillips-Perron test is that it is free from parametric errors. Phillips-Perron (PP) test allows the disturbances to be weakly dependent and heterogeneously distributed. In view of this, PP values have also been checked for stationarity. KPSS Test A major criticism of the ADF unit root testing procedure is that it cannot distinguish between unit root and near unit root processes especially when using short samples of data. This prompted the use of the KPSS test, where the null is of stationarity against the alternative of a unit root. This ensures that the alternative will be accepted (null rejected) only when there is strong evidence for (against) it [Kwiatkowski, Phillips, Schmidt. and Shin (1992)]. Co-integration Test Using non-stationary series, cointegration analysis has been used to examine whether there is any longrun equilibrium relationship. For instance, when non-stationary series are used in regression analysis, one as a dependent variable and the other as an independent variable, statistical inference becomes problematic [Granger and Newbold, 1974]. Cointegration analysis becomes important for the estimation of error correction models (ECM). The concept of error correction refers to the adjustment process between short-run disequilibrium and a desired long run position. As Engle and Granger (1987) have shown, if two variables are cointegrated, then there exists an error correction data generating mechanism, and vice versa. Since, two variables that are cointegrated, would on average, not drift apart over time, this concept provides insight into the long-run relationship between the two variables and testing for the cointegration between two variables. In the present case, Johansens Maximum Likelihood procedure for Cointegration has been applied. Johansen (1991) method can be illustrated by considering the following general autoregressive representation for the vector Y.
p

(2) where Yt is an n 1 vector of non stationary I(1) variables, A0 is an n 1 vector of constants, p is the number of lags, Aj is a (n x n) matrix of coefficients and t is assumed to be a ( n 1 ) vector of Gaussian error terms. In order to use Johansens test, the above vector autoregressive process can be reparametarized and turned into a vector error correction model of the form:
j 1

Yt

A0

AjYt

146
p 1

International Research Journal of Finance and Economics - Issue 14 (2008) Yt A0

j 1 p j i j 1 j

Yt

Yt

(3)

Where, Aj
p

and I
i j 1

Aj

is the difference operator, and I is an (n x n) identity matrix. The issue of potential co-integration is investigated by comparing both sides of equation (4). As Yt ~ I(1) , Yt ~ I(0) , so are Yt-j . This implies that the left-hand side of equation (4) is stationary. Since Yt-j is stationary, the right-hand side of equation (4) will also be stationary if Yt-p is stationary. The Johansen test centers on an examination of the matrix. The can be interpreted as a long run coefficient matrix, since in equilibrium, all the Yt-j will be zero, and setting the error terms, , to their expected value of zero will leave Yt-p = 0. The test for co-integration between the Ys is calculated by looking at the rank of the matrix via eigenvalues. The rank of a matrix is equal to the number of its characteristic roots (eigenvalues) that are different from zero. There are three possible cases to be considered: Rank ( ) = p and therefore vector Xt is stationary; Rank ( ) = 0 implying absence of any stationary long run relationship among the variables of Xt or Rank ( ) < p and equals to 0, the therefore r determines the number of cointegrating relationships. Thus, if the rank of is r matrix is null and equation (4) becomes the usual VAR model in first difference. If the rank of where r < n, then there exist r co-integrating relationships in the above model. In this case, the matrix ' and can be rewritten as = where are n r matrices of rank r. Here, is the matrix of co-integrating parameters and is the matrix of weights with which each co-integrating vector enters the above BVAR model. The test for the number of characteristic roots that are insignificantly different from unity can be conducted using the following two statistics, namely, the trace and the maximum eigenvalue test.
t p trace ( r )

T
j r 1

ln(1

) j ) r 1

(4)

and (5) is the estimated values of the characteristic roots (also called eigenvalue) obtained Where j matrix, T is the number of usable observations. r is the number of co-integrating from the estimated vectors. The trace test statistics test the null hypothesis that the number of distinct co-integrating vectors is less than or equal to r against the alternative hypothesis of more than r co-integrating relationships. From the above it is clear that trace equals zero when all j =0. The farther the estimated statistics. The characteristic roots are from zero, the more negative is ln(1- ) and larger the
max j trace

(r , r 1)

T ln(1

maximum eigenvalue statistics test the null hypothesis that the number of co-integrating vectors is less than or equal to r against the alternative of r +1 co-integrating vectors. Again, if the estimated value of the characteristic root is close to zero, max will be small (Enders, 1995; Madala and Kim, 1998).

International Research Journal of Finance and Economics - Issue 14 (2008)

147

Toda and Yamamoto (T-Y) Granger Causality Test The dynamic linkage is examined using the concept of Grangers causality test (1969, 1988). A time series xt Granger-causes another time series yt if series yt can be predicted with better accuracy by using past values of xt rather than by not doing so, other information is being identical. In other words, variable xt fails to Granger-cause yt if Pr( y t+m
t

) =Pr( y t+m
t

)
t

Where Pr( y t+m

) denotes conditional probability of yt , where

t

(6) is the set of all

information available at time t, and Pr( y t+m

) denotes conditional probability of yt obtained by

excluding all information on xt from yt . This set of information is depicted as t . The causal linkage between stock prices and macro economic aggregates in India are investigated by applying the technique of long run Granger non causality test developed by T-Y(1995). The selection of the VAR system requires an analysis of unit roots and cointegration which may cause inadequate results 1 (Blough, 1992). This can lead us to select an incorrect model for verifying the relations of causality, possibly causing a problem of over-rejection of non causal null hypothesis (Giles and Mirza, 1999). In this way, Toda and Yamamoto (1995), Dolado and Luketerpohl (1996) propose an applicable methodology independent of the integration or co-integration properties of the model. In this method a modified Wald Test is used to contrast the parameters of the VAR. An extended VAR model is used, whose order is determined by the number of optimal lag lengths in the system (k) and the maximum number of times one must differentiate the variables (d max ) . When a VAR (k d max ) is predicted (where d max is the maximum order of integration to occur in the system), this test displays asymptotic chi-square distribution, it is also shown that if variables are integrated of order d, the usual selection procedure is valid whenever k d. Toda and Yamamoto test has been used to capture long-run causality pattern of stock indices and the following specification has been used to estimate: Xt Yt
2 i 1 k d max 1 i 1 k d max 21 11

i Xt i Xt

k d max ( k d max ) 12 j 1 k d max ( k dmax ) 22 j 1

j Yt j Yt

( k dmax )

Xt

(7) (8)

( k dmax )

Yt

where Yt and xt are stationary random processes intended to capture other pertinent information not contained in lagged values of X t and Yt . The lag length k is decided by AIC in our study. The series Yt fails to Granger cause X t if 21(i)= 0 (i=1,2,3,.. k d max ). Variance decomposition The vector autoregression (VAR) by Sims (1980) has been estimated to capture short run causality between stock prices and key economic macro economic variables. Variance decomposition and impulse response function has been utilized for drawing inferences. Equation system is similar to the equations for granger causality assuming simultaneity among a set of variables and treating all the variables endogenous to system. VAR is commonly used for forecasting systems of interrelated time series and for analyzing the dynamic impact of random disturbances on the system of variables. The VAR approach sidesteps the need for structural modeling by treating every endogenous variable in the system as a function of the lagged values of all of the endogenous variables in the system. In the present study, BVAR model has been specified in first differences as given in equation 10 and 11.
1

12(j)= 0 (j=1,2,3, k d max ); and the series X t fails to cause Yt . If

The works of Bewley and Yang (1996) and Ho and Sorensen (1996) illustrate the problems associated with the utilization of the Johansen and Juselius test on the analysis of cointegration

148
k

International Research Journal of Finance and Economics - Issue 14 (2008)

k 11 i 1 k

Xt Yt

i i

Xt Xt

i j 1 k2 i j 1

12

j j

Yt Yt

Xt

(9) (10)

2 i 1

21

22

Yt

Where s are the stochastic error terms, called impulse response or innovations or shock in the language of VAR. Impulse response functions have been estimated to trace the effects of a shock to one endogenous variable on to the other variables in the BVAR. The impulse response functions can be used to produce the time path of the dependent variables in the BVAR, to shocks from all the explanatory variables. If the system of equations is stable any shock should decline to zero, an unstable system would produce an explosive time path. Variance decomposition (Choleski Decomposition) is the alternative way which separates the variation in an endogenous variable into the component shocks to the BVAR. Thus, the variance decomposition which provides information about the relative importance of each random innovation in affecting the variables in the BVAR has also been presented. In econometric literature, both impulse response functions and variance decomposition together are known as innovation accounting (Enders, 1995). For empirical investigation, the data set for the India consists of the quarterly NSE Nifty, BSE Sensex, Index of Industrial Production, Money Supply, Interest Rate, Foreign Direct Investment Inflows and Export Earnings for the period March, 1995 to March 2007. The Indian capital market is represented by the National Stock Exchange (NSE) and the Bombay Stock Exchange (BSE). The data for the NSE Nifty and BSE Sensex has been taken from their respective websites. The data for other macro key variables have taken from Ministry of Statistics and Programme Implementation, Govt. of India and Economic Intelligence Service, Centre for Monitoring Indian Economy Pvt. Ltd.

IV. Empirical Analysis

The descriptive statistics for all eight variables are calculated and presented in table 1. These variables are Export earnings, Money Supply, Exchange Rate, Interest Rate, Index of Industrial Production, Foreign Direct Investment, NSE Nifty and BSE Sensex. The skewness coefficient, in excess of unity is taken to be fairly extreme [Chou 1969]. High or low kurtosis value indicates extreme leptokurtic or extreme platykurtic [Parkinson 1987]. Generally values for zero skewness and kurtosis at 3 represents that the observed distribution is normally distributed. It is seen that the frequency distribution of the above mentioned variables are not normal. Jarque-Bera statistics also indicates that the frequency distribution of the underlying series does not fit normal distribution. Further, the standard deviation indicates that the Export earnings, Money Supply, FDI, NSE Nifty and BSE Sensex are relatively more volatile compared to Exchange Rate, Interest Rate and Index of Industrial Production.
Table 1:
Variables Mean Median Maximum Minimum Std. Dev. Skewness Kurtosis Jarque-Bera Probability

Descriptive Statistics
EXPORT 8.33 8.20 9.44 7.71 0.47 0.73 2.28 16.15 0.00 EXRATE 3.75 3.78 3.89 3.45 0.12 -1.02 2.85 25.17 0.00 IIP 5.12 5.09 5.65 4.73 0.21 0.28 2.30 4.90 0.09 INT 2.40 2.35 2.72 2.20 0.17 0.46 2.05 10.68 0.00 MS 14.06 14.09 15.01 13.18 0.52 -0.06 1.83 8.32 0.02 FDI 6.76 6.72 10.03 3.89 0.80 0.72 5.31 44.92 0.00 NIFTY 7.22 7.04 8.31 6.73 0.41 1.15 3.31 32.81 0.00 SENSEX 8.41 8.24 9.55 7.94 0.42 1.24 3.55 39.31 0.00

International Research Journal of Finance and Economics - Issue 14 (2008)

149

The first and simplest type of test one can apply to check for stationarity is to actually plot the time series and look for evidence of trend in mean, variance, autocorrelation and seasonality. If any such patterns are present then these are signs of non-stationarity. The eight time series displayed in figure-1 exhibit different such patterns. Export, IIP and Money Supply seems to exhibit a trend in the mean since it has a clear upward slope. In fact, sustained upward or downward sloping patterns (linear or non-linear) are signs of a non-constant mean. The time series on Exchange Rate, FDI, NSE Nifty and BSE sensex in the figure contains an obvious trend in both mean and variance. Along with a trend in the mean, the vertical fluctuation of the series also appears to differ greatly from one portion of the series to the other, indicating that the variance is not constant. Once again this is a sign of nonstationarity.

150

International Research Journal of Finance and Economics - Issue 14 (2008)

Figure 1: Dataset Graph
9 .6 9 .2 8 .8 8 .4 8 .0 7 .6 95 96 97 98 99 00 01 02 03 04 05 06 8 .4 8 .0 7 .6 7 .2 6 .8 6 .4 95 96 97 98 99 00 01 02 03 04 05 06

SENSEX

N IF T Y

9 .6 9 .2 8 .8 8 .4 8 .0

3 .9

3 .8

3 .7

3 .6

3 .5 3 .4

7 .6 95 96 97 98 99 00 01 02 03 04 05 06

95

96

97

98

99

00

01

02

03

04

05

06

E X R A TE

E X P ORT

15 .2

2.8
14 .8

2.7 2.6 2.5 2.4 2.3 2.2 2.1 95 96 97 98 99 00 01 INT 02 03 04 05 06

14 .4 14 .0 13 .6 13 .2 12 .8 95 96 97 98 99 00 01 MS 02 03 04 05 06

11
5 .8 5 .6 5 .4 5 .2 5 .0 4 .8

10 9 8 7 6 5 4

4 .6 95 96 97 98 99 00 01 IIP 02 03 04 05 06

3 95 96 97 98 99 00 01 FDI 02 03 04 05 06

International Research Journal of Finance and Economics - Issue 14 (2008)

Figure 2: Impulse Response Functions-Nifty and Other Macro Variables
R e s p o n s e o f D ( E X R A T E ) t o C h o le s k y O n e S . D . D ( N IF T Y ) In n o v a t i o n .0 0 3 .0 0 2 .0 0 1 .0 0 0 - .0 0 1 - .0 0 2 - .0 0 3 - .0 0 4 - .0 0 5 1 2 3 4 5 6 7 8 9 10

151

R e s p o n s e o f D (N IF T Y ) to C h o le s k y O n e S .D . D ( E X R A T E ) In n o va ti o n .0 0 5 .0 0 0 -.0 0 5 -.0 1 0 -.0 1 5 -.0 2 0 -.0 2 5 1 2 3 4 5 6 7 8 9 10

R e s p o n s e o f D ( E X P O R T ) to C h o le s k y O n e S . D . D ( N IF T Y ) In n o v a t i o n .0 6 .0 4 .0 2 .0 0 - .0 2 - .0 4 - .0 6 1 2 3 4 5 6 7 8 9 10
.0 2 0 .0 1 6 .0 1 2 .0 0 8 .0 0 4 .0 0 0 - .0 0 4 - .0 0 8 - .0 1 2 - .0 1 6 1 2 3 4 5 6 7 8 9 10 R e s p o n s e o f D ( N IF T Y ) to C h o le s k y O n e S . D . D ( E X P O R T ) In n o v a ti o n

R e s p o n s e o f D ( N IF T Y ) to C h o le s k y O n e S .D . D ( IIP ) In n o v a t i o n .0 2 4 .0 2 0 .0 1 6 .0 1 2 .0 0 8 .0 0 4 .0 0 0 - .0 0 4 - .0 0 8 - .0 1 2 1 2 3 4 5 6 7 8 9 10

R e s p o n s e o f D ( IIP ) t o C h o le s k y O n e S . D . D ( N IF T Y ) In n o v a t i o n .0 2

.0 1

.0 0

-.0 1

-.0 2 1 2 3 4 5 6 7 8 9 10

R e s p o n s e o f D (M S ) to C h o le s k y O n e S . D . D ( N I F T Y ) I n n o v a ti o n .0 0 15 .0 0 10 .0 0 05 .0 0 00 -.0005 -.0010 -.0015 -.0020 1 2 3 4 5 6 7 8 9 10 .0 2 5 .0 2 0 .0 1 5 .0 1 0 .0 0 5 .0 0 0 -.0 0 5 -.0 1 0 1 2

R e s p o ns e o f D (N IF TY) to C ho le s k y O ne S .D . D (M S ) Inno va ti o n

10

152

International Research Journal of Finance and Economics - Issue 14 (2008)

Response of D(NIFTY) to Cholesky One S.D. D(INT) Innovation .03 .02 .01 .00 -.01 -.02 -.03 1 2 3 4 5 6 7 8 9 10
.004 .003 .002 .001 .000 -.001 -.002 -.003 -.004 -.005 1 2 3 4 5 6 7 8 9 10 Response of D(INT) to Cholesky One S.D. D(NIFTY) Innovation

R e s p o n s e o f D (N IF T Y) to C h o le s k y O n e S .D . D ( F D I) In n o v a tio n .0 3 .0 2 .1 .0 1 .0 .0 0 -.0 1 -.0 2 1 2 3 4 5 6 7 8 9 10 -.1 .2

R e s p o ns e o f D (F D I) to C ho le s k y O ne S .D . D (N IF TY) Inno va tio n

-.2 1 2 3 4 5 6 7 8 9 10

R e s p o n s e o f D ( N IF T Y ) to C h o le s k y O n e S .D . D ( S E N S E X ) In n o v a ti o n .1 0 .0 8 .0 6 .0 4 .0 2 .0 0

R e s p o n s e o f D ( S E N S E X ) t o C h o le s k y O n e S .D . D ( N IF T Y ) In n o v a ti o n .0 3 .0 2 .0 1 .0 0 - .0 1 - .0 2

- .0 2 - .0 4 1 2 3 4 5 6 7 8 9 10 - .0 3 1 2 3 4 5 6 7 8 9 10

International Research Journal of Finance and Economics - Issue 14 (2008)

Figure 3: Impulse Response Functions-Sensex and Other Macro Variables
Re sp o nse o f D (E X RA TE ) to C ho le sky O ne S .D . D (S E NS E X ) Inno va tio n .0 01 .0 00 -.00 1 -.00 2 -.00 3 -.00 4 -.00 5 -.00 6 1 2 3 4 5 6 7 8 9 10
Re sp o nse o f D (S E N S E X ) to C ho le s ky O ne S .D . D (E X R A TE ) Inno va tio n .0 1 0 .0 0 5 .0 0 0 -.0 0 5 -.0 1 0 -.0 1 5 -.0 2 0 -.0 2 5 1 2 3 4 5 6 7 8 9

153

10

R e sp o nse o f D (E X P O R T) to C hole s ky O ne S .D . D (S E N S E X ) Inno va tio n

R e s p o ns e o f D (S E N S E X ) to C ho le s k y O ne S .D . D (E X P O R T) Inno va ti o n .0 2 0 .0 1 5 .0 1 0 .0 0 5 .0 0 0 -.0 0 5 -.0 1 0 -.0 1 5 -.0 2 0 1 2 3 4 5 6 7 8 9 10

.04 .03 .02 .01 .00 -.01 -.02 -.03 1 2 3 4 5 6 7 8 9 10

Response of D (IIP ) to C holesky One S .D . D (S E NS E X ) Innovation .015

.0 2 0

R e s p o n s e o f D ( S E N S E X ) to C h o le s k y O n e S .D . D ( IIP ) In n o v a ti o n

.010 .005 .000 -.005 -.010 -.015 -.020 1 2 3 4 5 6 7 8 9 10

.0 1 5 .0 1 0 .0 0 5 .0 0 0 - .0 0 5 - .0 1 0 - .0 1 5 1 2 3 4 5 6 7 8 9 10

R e s p o ns e o f D ( IN T ) to C ho le s k y O ne S .D . D (S E N S E X ) Inn o v a ti o n .0 0 4
.0 3 .0 2

R e s p o n s e o f D ( S E N S E X ) to C h o le s k y O n e S .D . D ( IN T ) In n o v a ti o n

.0 0 2
.0 1

.0 0 0 -.0 0 2

.0 0 - .0 1 - .0 2 - .0 3

-.0 0 4 -.0 0 6 1 2 3 4 5 6 7 8 9 10

10

154
.003 .002 .001 .000 -.001 -.002 -.003 1 2 3 4 5

International Research Journal of Finance and Economics - Issue 14 (2008)

Response of D (MS ) to C holesky One S .D . D (S E NS E X ) Innovation .010 .005 .000 -.005 -.010 -.015 -.020 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Response of D (SE NS E X ) to C holesky One S .D . D (MS ) Inno vation

Response of D(SENSEX) to Cholesky One S.D. D(FDI) Innovation .03 .02 .01
.0 .2

Response of D(FDI) to Cholesky One S.D. D (SENSEX) Innovation

.1

.00
-.1

-.01 -.02 1 2 3 4 5 6 7 8 9 10
-.2 1 2 3 4 5 6 7 8 9 10

As indicated by visual inspection, formal test for stationarity is essential to opt for appropriate methodological framework. As a first step, we tested the eight variables (Exports, Exchange Rate, Index of Industrial Production, Interest Rate, Money Supply, FDI, Nifty and Sensex) for stationarity by applying ADF, PP unit root test and KPSS stationarity test. ADF, PP and KPSS statistics are given in Table-2. On the basis of ADF statistics and PP test, all the series are found to be non-stationary at levels except exchange rate. In the series of exchange rate, ADF and PP test statistics rejects null hypotheses of unit root in case of levels at 5 percent and 6 percent level of significance. Further, ADF statistics rejects null hypotheses of unit root in case of first differences of money supply only at 10 percent. Finally, KPSS test is applied which has null of stationarity. In this case, all variables including exchange rate are non stationary in levels and stationary in first differences. Assuming all the variables are non-stationary at levels and stationary at first differences on the basis of KPSS test and visual inspections, Johansens approach of cointegration, Toda and Yamamoto Granger causality test and VAR modeling for variance decomposition/impulse response functions have been applied.

International Research Journal of Finance and Economics - Issue 14 (2008)

Table 2: Unit Root Test
Null Hypothesis: Variable Null Hypothesis: Variable is Non-stationary is Non-stationary Augmented Dickey-Fuller Phillips-Perron Test test statistic Statistics Level First Difference Level First Difference 2.43 -2.93* -0.07 -27.40* -2.88** -9.04** -2.86*** -8.98* 3.13 -2.71** 1.01 -23.69* 0.91 -9.99* 0.09 -55.39* -1.10 -9.06* -1.18 -8.85* -0.03 -2.68*** 0.93 -10.59* 0.14 -13.10* 0.19 -13.07* 0.37 -12.56* 0.33 -12.56* Asymptotic critical values*: -3.48 -3.47 -2.88 -2.88 -2.57 -2.57

155

Variables Exports Exchange Rate Index of Industrial Production FDI Interest Rate Money Supply Nifty Sensex 1% level 5% level 10% level

Null Hypothesis: Variable is stationary Kwiatkowski-PhillipsSchmidt-Shin test statistic Level First Difference 1.31* 0.39 1.02* 0.65 1.41* 0.37 0.68** 0.14 1.32* 0.15 1.42* 0.15 1.06* 0.29 0.95* 0.31 0.74 0.46 0.35

* implies significant at 1% level, ** implies significant at 5% level and *** implies significant at 10% level

To explore whether there is any long-run relationship between Indian stock markets and macro economic variables such as exports, exchange rate, index of industrial production, foreign direct investment, interest rate and money supply, Johansens cointegration test has been applied. The number of lags in cointegartion analysis is chosen on the basis of Akaike Information Criteria. Before discussing the results, it is important to discuss what it implies when two variables are cointegrated and when they are not. When two variables are cointegrated, it implies that the two time series cannot wander off in opposite directions for very long without coming back to a mean distance eventually. But it doesnt mean that on a daily basis the two series have to move in synchrony at all. When two series are not cointegrated it implies that the two time series can wander off in opposite directions for very long without coming back to a mean distance eventually. Results indicate that NSE Nifty-Index of Industrial production and Money Supply- NSE Nifty may be cointegrated in the long run as the results vary depending on the varying assumption about trend and intercept. However, FDI- NSE Nifty is cointegrated in the long run under all assumptions. In case of Exchange Rate-NSE Nifty; Export- NSE Nifty and Interest Rate- NSE Nifty, there is no evidence of co-integration. Similar test was conducted between Sensex and above referred macro variables. The results remain the same except that the study could not reveal any long run relationship between FDI and Sensex as reported in case of NSE and Nifty case (see Table-3 and 4).

156
Table 3:

International Research Journal of Finance and Economics - Issue 14 (2008)

Johansen Co-Integration Test: Nifty and Other Macro Variables (Number of Cointegrating Relations by Model)
None No Intercept No Trend 0 0 2 0 0 0 0 0 1 1 1 1 0 0 None Linear Intercept Intercept No Trend No Trend Nifty- Exchange Rate(3) 0 0 0 0 Nifty- Exports(3) 0 0 0 0 Nifty- Index of Industrial Production(2) 0 0 0 0 Nifty- Interest Rate(4) 0 0 0 0 Nifty- Money Supply(1) 1 0 1 0 Nifty- Foreign Direct Investment(2) 1 1 1 1 Nifty-Sensex(3) 0 0 0 0 Linear Intercept Trend 0 0 0 0 1 1 0 0 0 0 1 1 0 0 Quadratic Intercept Trend 0 0 0 0 1 1 0 0 0 0 1 1 0 0

Data Trend: Test Type Trace Max-Eig Trace Max-Eig Trace Max-Eig Trace Max-Eig Trace Max-Eig Trace Max-Eig Trace Max-Eig

Selected (0.05 level*) Appropriate lag is given in parantheses on the basis of AIC

To test long run direction of causality, Toda and Yamamoto Granger causality test was conducted between Nifty and other variables. The results are presented in Table-4. In case of exchange rate-NSE Nifty and IIP-NSE Nifty, there is no causality in either direction. The study reveals that export performance does not have any effect on NSE Nifty while NSE Nifty affects export flows from India. Further, results reveal that Interest Rate causes NSE Nifty while there is no causality from NSE Nifty to Interest Rate. In case of Money Supply and NSE Nifty, there is no direct causal links. Money supply might be affecting Nifty through its effect on interest rate. In case of FDI-NSE Nifty, FDI does cause Nifty movement while movement in NSE Nifty does not have significant effect on IIP. This links may be the outcome of positive effect of FDI inflows on expectations regarding growth, earning and profit prospect of the sector. In case of Nifty and Sensex, Sensex causes Nifty while movement in NSE Nifty does not have significant effect on Sensex.

International Research Journal of Finance and Economics - Issue 14 (2008)

Table 4:

157

Johansen Co-Integration Test: Sensex and Other Macro Variables (Number of Cointegrating Relations by Model)
None No Intercept No Trend 0 0 1 0 1 0 0 0 1 1 0 0 None Linear Intercept Intercept No Trend No Trend Sensex- Exchange Rate(2) 1 0 1 0 Sensex- Exports(3) 0 0 0 0 Sensex- Index of Industrial Production(3) 0 0 0 0 Sensex- Interest Rate(4) 0 0 0 0 Sensex- Money Supply(1) 1 0 1 0 Nifty- Foreign Direct Investment(4) 0 0 0 0 Linear Intercept Trend 0 0 0 0 1 1 0 0 0 0 0 0 Quadratic Intercept Trend 0 0 0 0 1 1 0 0 0 0 0 0

Data Trend: Test Type Trace Max-Eig Trace Max-Eig Trace Max-Eig Trace Max-Eig Trace Max-Eig Trace Max-Eig

Selected (0.05 level*) Appropriate lag is given in parantheses on the basis of AIC

Similar Toda and Yamamoto test for causality were conducted between BSE Sensex and other macro economic variables. The results are presented in Table-5. The results reveal that exchange rate does not influence the BSE Sensex while movement in BSE Sensex does cause change in exchange rate. There is no evidence that export performance does have any effect on BSE Sensex while BSE Sensex causes export flows from India. It is found that Index of Industrial production (IIP) does not have any influence on Sensex movement while movement in BSE Sensex has significant effect on IIP. Results reveal that Interest Rate causes BSE Sensex while there is no causality from BSE Sensex to Interest Rate. In case of BSE Sensex -Money Supply and BSE Sensex -FDI, there is no causal link observed in ether direction.
Table 5:
Cause Exchange Rate Nifty Export Nifty Index Of Industrial Production(IIP) Nifty FDI Nifty Interest Rate Nifty Money Supply Nifty Nifty Sensex

T&Y(1995) Test of Granger Causality: Nifty and Other Macro Variables.

Effect Nifty Exchange Rate Nifty Export Nifty Index of Industrial Production(IIP) Nifty FDI Nifty Interest Rate Nifty Money Supply Sensex Nifty Wald Test: Chisquare Values 5.24(3+1) 0.92(3+1) 0.56(3+1) 7.38**(3+1) 3.66(2+1) 2.52(2+1) 5.01***(2+1) 0.92(2+1) 9.54*(4+1) 4.88(4+1) 0.94(1+1) 0.03(1+1) 1.86(3+1) 1382.92*(3+1) Causality Inference Exchange Rate does not causes Nifty Nifty does not causes Exchange Rate Export does not causes Nifty Nifty causes Export IIP does not cause Nifty Nifty does not cause IIP FDI cause Nifty Nifty does not cause FDI Interest Rate causes Nifty Nifty does not cause Interest Rate Money Supply does not cause Nifty Nifty causes Money Supply Nifty does not cause Sensex Sensex causes Nifty Model VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax)

* implies significant at 1% level, ** implies significant at 5% level and *** implies significant at 10% level.

158

International Research Journal of Finance and Economics - Issue 14 (2008)

In the context of varying causal links of both stock indices with macro variables, Simss VAR were applied and short run causal links were explored using Variance decomposition and Impulse response functions. Variance decomposition determines how much of the k-step ahead forecast error variance of given variable is explained by innovations to each explanatory variable. In practice, it is usually observed that own series shocks most of the (forecast) error variance of the series in the VAR. Variance decomposition separates the variation in an endogenous variable into the component shocks to the VAR and provides information about the relative importance of each random innovation in affecting the variables in the VAR. The variance decomposition results at the end of 10 periods are shown in Table 7. The columns provide the percentage of the forecast variance due to each innovation in bivariate VAR (BVAR) framework, with each row adding up to 100. In case of bivariate modeling of Nifty and exchange rate, NSE Nifty explains nearly 97% of its own forecast error variance while Exchange rate explains only 3 percent of NSE Nifty. Exchange Rate explains over 91% of its own variance and Nifty explains remaining 9 percent of variance of exchange rate. It implies that NSE Nifty causes Exchange Rate. Variance decomposition in case of Nifty-Export BVAR Model indicates that NSE Nifty explains almost 100 percent of its variance while NSE nifty explains about 7 percent of variance of exports. It is an indication that NSE Nifty causes exports. In case of Nifty and IIP BVAR, Nifty explains 23 percent of variance of IIP while IIP explains about 20 percent of variance in NSE Nifty. In this case, Nifty leads IIP. In case of Nifty-Interest Rate BVAR, results indicate that interest rate explains 7.29 percent variance in NSE Nifty while NSE Nifty explains 3.72 percent of Interest rate, thus Interest Rate affects NSE Nifty. In case of Nifty and FDI BVAR, FDI explains more than 4 percent of variance of Nifty while Nifty explains only 2 percent of FDI variance. FDI seems to have some lead to NSE Nifty. In case of Nifty and Sensex BVAR, Sensex explains 90 percent of variance of Nifty while Nifty explains about 3 percent of variance in Sensex. In this case, it is quite clear that Sensex causes Nifty. Money supply only explains 2 percent of NSE Nifty while 5 percent of variance of Money supply is explained by NSE Nifty. Direction of effect seems from NSE Nifty to Money Supply.
Table 6:
Cause Exchange Rate Sensex Export Sensex Index Of Industrial Production(IIP) Sensex FDI Sensex Interest Rate Sensex Money Supply Sensex

T&Y(1995) Test of Granger Causality:Sensex and Other Macro Variables

Effect Sensex Exchange Rate Sensex Export Sensex Index of Industrial Production(IIP) Sensex FDI Sensex Interest Rate Sensex Money Supply Wald Test: Chisquare Values 1.92(2+1) 8.35*(2+1) 1.38(3+1) 6.24**(3+1) 3.06(3+1) 8.14*(3+1) 4.21(4+1) 4.58(4+1) 10.76*(4+1) 5.37(4+1) 0.37(1+1) 1.00(1+1) Causality Inference Excahnge Rate does not causes Sensex Sensex causes Exchange Rate Export does not causes Sensex Sensex causes Export IIP does not cause Sensex Sensex cause IIP FDI does not cause Sensex Sensex does not cause FDI Interest Rate causes Sensex Sensex does not cause Interest Rate Money Supply does not cause Sensex Sensex causes Money Supply Model VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax) VAR(k +dmax)

* implies significant at 1% level, ** implies significant at 5% level and *** implies significant at 10% level.

International Research Journal of Finance and Economics - Issue 14 (2008)

Table 7:

159

VARIANCE DECOMPOSITION OF NIFTY AND OTHER VARIABLES (Bivariate VAR Framework)

Lags: 2 5 10 2 5 10 Lags: 2 5 10 2 5 10 Lags: 2 5 10 2 5 10 Lags: 2 5 10 2 5 10 Lags: 2 5 10 2 5 10 Lags: 2 5 10 2 5 10 Lags: 2 5 10 2 5 10 Nifty 98.58 96.06 96.04 4.99 5.16 5.16 Nifty 99.85 99.46 99.45 2.43 8.54 8.63 Nifty 98.31 98.35 98.35 10.79 12.59 12.63 Nifty 99.99 92.94 92.67 0.23 3.49 3.67 Nifty 99.43 99.43 99.43 0.12 0.12 0.12 Nifty 96.46 95.77 95.70 0.65 1.91 1.92 Nifty 9.98 9.71 9.69 3.21 3.10 3.10 Exchange Rate 1.41 3.93 3.95 95.00 94.83 94.83 Export 0.14 0.53 0.54 97.56 91.45 91.36 IIP 1.68 1.64 1.64 89.20 87.40 87.36 Interest Rate 0.01 7.059 7.32 99.76 96.50 96.32 Money Supply 0.56 0.56 0.56 99.87 99.87 99.87 FDI 3.53 4.22 4.29 99.34 98.08 98.08 Sensex 90.01 90.28 90.30 96.78 96.89 96.89

Variance Decomposition of Nifty Exchange Rate

Nifty Export

Nifty IIP

Nifty Interest Rate

Nifty Money Supply

Nifty FDI

Nifty Sensex

Similarly, the variance decomposition results at the end of 10 periods for BSE Sensex in bivariate framework are shown in Table 8. In case of bivariate modeling of BSE Sensex and exchange rate, BSE Sensex explains nearly 98% of its own forecast error variance while Exchange rate explains only 2 percent of BSE Sensex. Results indicate Exchange Rate explains over 91% of its own variance and Sensex explains remaining 9 percent of variance of exchange rate. It implies that BSE Sensex

160

International Research Journal of Finance and Economics - Issue 14 (2008)

causes Exchange Rate. Results indicate that BSE Sensex explains almost 99 percent of its variance and about 3 percent of variance of exports while export explains about 1 percent of variance in BSE Sensex and 97 percent its own variance. BSE Sensex seems to cause export flows. In case of Sensex and IIP BVAR, Sensex explains 5 percent of variance of IIP while IIP explains about 1 percent of variance in BSE Sensex. In this case, Sensex causes IIP. Sensex-Interest BVAR Rate results indicate that interest rate explains 6.41 percent variance in BSE Sensex while BSE Sensex explains only 0.12 percent of Interest rate, thus Interest Rate causes BSE Sensex. Money supply only explains 2.44 percent of BSE Sensex while 6.90 percent of variance of Money supply is explained by BSE Sensex in case of SensexMoney Supply BVAR model. Direction of effect seems from BSE Sensex to Money Supply. It is difficult to infer in case of Sensex-FDI BVAR, as small proportion of variance of each other are explained in this case. FDI explains approximately 3 percent of Sensex variance while BSE sensex explains 2 percent of FDI variance. However, FDI may be causing movement in BSE Sensex in this case.
Table 8: Variance Decomposition of Sensex and other variables
At Time Period 2 5 10 2 5 10 2 5 10 2 5 10 2 5 10 2 5 10 2 5 10 2 5 10 2 5 10 2 5 10 Lags: 2 5 10 2 5 10 Sensex 98.71 98.40 98.40 8.37 9.42 9.42 Sensex 99.78 99.04 99.04 0.88 2.88 3.08 Sensex 99.19 98.45 98.37 2.14 5.08 5.06 Sensex 94.85 92.53 92.25 1.43 3.46 3.80 Sensex 99.65 99.65 99.65 1.65 1.65 1.65 Sensex 99.95 97.14 96.97 1.16 1.95 2.09 Exchange Rate 1.28 1.59 1.59 91.62 90.57 90.57 Export 0.21 0.95 0.95 99.11 97.11 96.91 IIP 0.80 1.54 1.62 97.85 94.91 94.93 Interest Rate 5.14 7.46 7.74 98.56 96.53 96.19 Money Supply 0.34 0.34 0.34 98.34 98.34 98.34 FDI 0.04 2.85 3.02 98.83 98.04 97.90

Variance Decomposition of Sensex Exchange Rate

Sensex Export

Sensex IIP

Sensex Interest Rate

Sensex Money Supply

Sensex FDI

To investigate dynamic responses further between the variables, Impulse Response of the VAR system has also been estimated. An Impulse Response function traces the effect of a one-time shock to

International Research Journal of Finance and Economics - Issue 14 (2008)

161

one of the innovations on current and future values of the endogenous variables. So, for each variable from each equation separately, a unit shock is applied to the error, and the effects upon the VAR system over time is noted. A shock to the i-th variable not only directly affects the i-th variable but is also transmitted to all of the other endogenous variables through the dynamic (lag) structure of the VAR. Thus, if there are g variables in a system, total of g2 impulse responses could be generated. In our case, we have thirteen BVAR pairs of variables; technically four graphs of Impulse Responses (IR) for each BVAR pair are possible in each period totaling fifty two. However, we have presented only twenty-six IR graph for the sake of brevity because IR graph representing direct affects on the i-th variable have not been presented as it does not reveal any causal pattern of response in theses cases. As indicated by variance decomposition, similar pattern of causality is also observed graphically using impulse response functions. Further, Impulse Response Function indicates that the BVAR system is stable.

6. Concluding Remarks
The purpose of the present study is to explore the causal relationships between stock prices and the key macro variables representing real and financial sector of the Indian economy. These variables are the index of industrial production, exports, foreign direct investment, money supply, exchange rate, interest rate, NSE Nifty and BSE Sensex. The present analysis is based on quarterly data from March, 1995 to March, 2007. Johansens approach of cointegration and T-Y Granger causality test have been applied to explore the long-run causal relationships while BVAR modeling for variance decomposition and impulse response functions has been applied to examine short run causal relationships. Co-integration regressions indicate the presence of a long run relationship between stock prices and FDI, stock prices and MS and stock prices and IIP. However, except for interest rate and exports, the pattern of cointegration and long-run Granger causality in each market indicates differential pattern. In NSE, movement in NSE does not have effect on exchange rate and IIP while movement in BSE Sensex seems to cause these variables. In this context, it is important to highlight that the composition of NSE Nifty and BSE Sensex vary from each other. Nifty is a benchmark index that is composed of 50 stocks i.e. shares of 50 companies. Sensex, on the other hand, is an index that is composed of thirty stocks i.e. shares of 30 companies. The rate of change in variation may happen because every stock in an index has some index-weightage and it is varying in each index. In case of short run, there is no differential impact of causal pattern as indicated by variance decomposition and impulse response functions. In this case, the results reveal that NSE Nifty causes exchange rate, exports, IIP and money supply while interest rate and FDI causes NSE Nifty. Broadly, these patterns are valid also in the case of BSE Sensex. The study reveals that movement in stock prices causes movement in IIP. It clearly implies that stock prices leads real economic activity. It reveals that growth rate of real sector is factored in the movement in stock prices. One possible explanation is that major players such as FIIs and mutual funds are well aware in advance of business plan of major industrial houses. Another explanation is the higher demand of stocks created by positive growth expectations. Higher demand has been created by increased due to increased participation of domestic retail investors and trader directly or through mutual fund as the expected returns in alternative investment opportunities such as bank deposits, deposits in post offices, etc are less than the expected return in stocks. Further, FIIs inflows has increased in recent years due to expected higher growth of rate of return in Indian compared to USA, EU and other developed economies. This has been possible due to change in legal and institutional trading environment. This study also reveals that FDI does cause movement in stock prices while movement in stock prices does not have significant effect on FDI. This links may be the outcome of positive effect of FDI inflows on expectations regarding growth, earning and profit prospect of the sector. Results indicate that the movements in stock prices seem to affect export flows, possibly through its effect on exchange rate. It is to be noted that the movement in Sensex and Nifty are causing changes in exchange rate at least in the short run. One possible inference may be drawn from the

162

International Research Journal of Finance and Economics - Issue 14 (2008)

present study is that the stock market movement in India seems to be driven by the performance of industry in domestic market rather than in export market. It seems to be intuitively correct as the export segment of major listed companies is small barring IT and textile sector. The study indicates that exchange rate does not influence the BSE Sensex and NSE Nifty while movement in BSE Sensex and NSE Nifty does cause change in exchange rate. It indicates towards large FII inflows in recent years. It is causing serious problem to the real sector of the economy. As of October, the rupee had appreciated close to 13 per cent on a year-on-year basis against the greenback. Nearly 7.5 per cent of this appreciation came in the months of April and May. In recent months, net FII inflows have picked up around $5.7 billion in October itself. The recent cuts in the Fed rate are likely to provide a further impetus to the inflows and put more pressure on the rupee. One of the major worries of rupee appreciation is the hit exports are expected to take. The relationship between money supply and stock prices has been widely studied because of the belief that money supply changes have important direct effects through portfolio changes, and indirect effects through their effect on real economic activity, which in turn postulated to be the fundamental determinants of stock prices. However, the study indicates that money supply does not affect movement in stock market indices while interest rate causes change in stock indices. The possible explanation may be the lack of understanding on the part of players in the stock market about the complex linkage of money supply with cost and profit margin of industry. Further, major player are either not able to predict the possible effect of money supply and monetary measures related to money supply taken by RBI or they could not form their expectation on the basis of changes in money supply. Given the demand for money, there is a direct inverse relationship between money supply and interest rate. Probably, stock market reacts to money supply only if it causes changes in interest rate. It may be inferred that monetary policy measures affecting interest rate are critical in influencing stock market in India.

International Research Journal of Finance and Economics - Issue 14 (2008)

163

References
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] Abdalla, I. S. A. and V. Murinde (1997), Exchange Rate and Stock Price Interactions in Emerging Financial Markets: Evidence on India, Korea, Pakistan, and Philippines, Applied Financial Economics 7, 25-35. Abdalla, I. S. A. and V. Murinde (1996) Exchange rate and stock prices interactions in emerging financial markets: Evidence on India, Korea, Pakistan and Philippines. Applied Financial Economics, 7, 25-35. Ando, A. and F. Modigliani (1963), The Life Cycle Hypothesis of Saving: Aggregate Implications and tests, American Economic Review, Vol. 53, No. 1. Ajayi, R. A. and M. Mougoue (1996), On the Dynamic Relation between Stock Prices and Exchange Rates, Journal of Financial Research 19, 193-207. Bahmani-Oskooee, M. and A. Sohrabian (1992). Stock Prices and the Effective Exchange Rate of the Dollar, Applied Economics, 24, 4, 459-464. Baumol, W. (1965), Stock Market and Economic Efficiency, Fordham University Press, New York. Bernanke, B.S. and K.N. Kuttner (2005), What Explains the Stock Markets Reaction to Federal Reserve Policy,Journal of finance, 60, 1221-1257. Bewley, R., and Orden, D.(1994), Alternative methods of estimating long run responses with applications to Australian import demand, Econometric Reviews, 13, 179-204. Blough, S. R. (1992), The Relationship Between Power and Level for Generic Unit Root Tests in Finite Samples, Journal of Applied Econometrics, 7, 295308. Bhattacharya B and Mukherjee J. (2002), Causal relationship between stock market and exchange rate, foreign exchange reserves and value of trade balance: A case study for India,. www.igidr.ac.in Bodie, Zvi. (1976), Common Stocks As A Hedge Against Inflation, Journal of Finance, 31(2), 459-470. Bosworth, B.(1975), "The Stock Market and the Economy", Brookings Papers on Economic Activity, Vol. 2. Chen, N, R. Roll, and S. Ross (1986), Economic forces and the stock market, Journal of Business 59, 383-403. Chou, Y L (1969): Statistical Analysts, Holt Rinehart and Winston, London. Cutler, D. M., J. M. Poterba, and L. H. Summers (1989), What moves stock prices?, Journal of Portfolio Management, 15, 4-12. Dickey, D. A. and W. A. Fuller (1979), Distribution of the Estimators for Autoregressive Time Series with a Unit Root, Journal of the American Statistical Association 74, 427-431. Dickey, D. A. and W. A. Fuller (1981), Likelihood Ratio Statistics for Autoregressive Time Series with a Unit Root, Econometrica 49, 1057-1072. Dolado, J.J. and Lutkepohl, H. (1996), Making Wald Test Work for Cointegrated VAR Systems, Econometrics Reviews, 15, 369-386. Dornbusch, R. and S. Fischer (1980), Exchange Rates and Current Account, American Economic Review, 70, 960-971. Enders, W. (1995), Applied Economic Time Series, Wiley, New York. Engle, R. F. and C. W. J. Granger (1987), Co-integration and error-correction: Representation, estimation and testing, Econometrica, 55, 251276. Fama, E. F. (1981), Stock Returns, Real Activity, Inflation, and Money, American Economic Review, 71(4), 545-565. Friedman, M. (1988), Money and the Stock Market, Journal of Political Economy, 96, 2, 221-245. Geske, R. and R. Roll (1983), The Fiscal and Monetary Linkage between Stock Returns and Inflation, Journal of Finance, 38(1), 1-34.

164 [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42]

International Research Journal of Finance and Economics - Issue 14 (2008) Giles, J.A. and Mirza, S. (1999), Some pretesting issues on testing for Granger noncausality, Econometrics Working Paper EWP 9914, Department of Economics, University of Victoria. Granger, C. W. J. (1988), Some Recent Developments in the Concept of Causality, Journal of Econometrics, 39, 199-211. Granger, C. W. J. and Newbold, P. (1974), Spurious Regression in Econometrics Journal of Econometrics, 2, 111-120. Granger, C. W. J. and Newbold, P. (1974), Spurious Regression in Econometrics Journal of Econometrics, 2, 111-120. Gujarati, D.N.(2003), Basic Econometrics, McGraw Hill, India. Ho, M. S., and B. E. Sorensen (1996), Finding cointegration rank in high dimension systems using the Johansen test. An illustration uses data based Monte Carlo simulations, Review of Economics and Statistics, 78, 726- 732. Johansen, S. (1991), Estimation and hypothesis testing of cointegration vector in Gaussian vector autoregressive models, Econometrica, 59, 1551-1581. Kwiatkowski, D., Philips, P.C.B., Schmidt, P., Shin, Y., (1992), Testing the Null Hypothesis of Stationary against the Alternative of a Unit Root. How Sure Are We that Economic Time Series Have a Unit Root?, Journal of Econometrics, 54, pp.159-178. Ma, C. K. and G. W. Kao (1990), On Exchange Rate Changes and Stock Price Reactions, Journal of Business Finance and Accounting, 17, 441-449. Maddala G.S. and I.M. Kim (1998), Unit Roots, Cointegration and Structural Change, Cambridge University Press. Mukherjee, T.K. and A. Naka, (1995), Dynamic relations between macroeconomic variables and the Japanese stock market: An application of a vector error correction model, The Journal of Financial Research, 2, 223-237. Pan, Fok & Lui (1999), Dynamic Linkages Between Exchange Rates and Stock Prices: Evidence from Pacific Rim Countries, Shippensburg University Working Papers. Parkinson, J M (1987): The EMH and the CAPM on the Nairobi Stock Exchange, East African Economic Review, 13, 105-110. Pearce, D. K. and V. V. Roley (1983), The Reaction of Stock Prices to Unanticipated Changes in Money: A Note, Journal of Finance 38, 1323-1333. Pearce, D.K. (1985), Stock Prices and Economic News, Journal of Business, 58(1), 49-67. Phillips, P. C. and Perron, P. (1988), Testing for a Unit Root in Time Series Regression Biometrika, 75, 335-346. Sims, C. A. (1980), Macroeconomics and Reality, Econometrica, 48(1): 1-48. Toda, H.Y. and T. Yamamoto (1995), Statistical inferences in vector autoregressions with possibly integrated processes, Journal of Econometrics, 66, 225-50.