Professional Documents
Culture Documents
Dynamic Relation Between Exchange Rate and Stock Prices - A Case For India Golaka C. Nath and G. P. Samanta
Dynamic Relation Between Exchange Rate and Stock Prices - A Case For India Golaka C. Nath and G. P. Samanta
Introduction
The Asian crisis of 1997-98 has made a strong pitch for dynamic linkage between stock prices and
exchange rates. During the crisis period, the world has noticed that the emerging markets collapsed
due to substantial depreciation of exchange rates (in terms of US$) as well as dramatic fall in the
stock prices. This has become important again from the view point of large cross border movement
of funds due to portfolio investment and not due to actual trade flows, though trade flows have
some impact on stock prices of the companies whose main sources of revenue comes from foreign
exchange.
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 country’s 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.
Although theories suggest causal relations between stock prices and exchange rates, existing
evidence on a micro level provides mixed results. Jorion (1990, 1991), Bodnar and Gentry (1993),
and Bartov and Bodnar (1994) all fail to find a significant relation between simultaneous dollar
movements and stock returns for U.S. firms. He and Ng (1998) find that only about 25 percent of
their sample of 171 Japanese multinationals has significant exchange rate exposure on stock returns.
Griffin and Stulz’s (2001) empirical results show that weekly exchange rate shocks have a negligible
impact on the value of industry indexes across the world. However, Chamberlain, Howe, and Popper
(1997) find that the U.S. banking stock returns are very sensitive to exchange rate movements, but
∗
Manager, NSE
∗∗
Assistant Adviser, Reserve Bank of India. The views expressed and the approach suggested are of the
authors and not necessarily of the organizations to which they belong.
not for Japanese banking firms. While such findings are different from those reported in prior
research, Chamberlain et al. attributed the contrast to the use of daily data in their study instead of
monthly data as used in most prior studies.
On a macro level, the empirical research documents 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.
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 country’s
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.
In the context of Indian economy, however, study in the similar direction is not available,
though the issue is gaining importance in recent years. Like other Asian emerging economies, Indian
equity market has continued to grow and has seen the relaxation of foreign investment restrictions
primarily through country deregulation. During the 1990s, India has initiated the financial sector
reforms by way of adopting international practices in its financial market. Parallel to this, the issuance
of American Depository Receipts (ADR’s) or General Depository Receipts (GDR’s) has facilitated
the trade of foreign securities on the NYSE, NASDAQ or on non-American exchanges. Over the
years, Indian Rupee is slowly moving towards full convertibility, which has also had an impact in the
Indian capital market as international investors have invested about US $15 billion in Indian capital
market. The two-way fungibility of ADRs/GDRs allowed by RBI has also possibly strengthened the
linkages between the stock and foreign exchange markets in India. In this background, this study
aims at examining the dynamic linkages between foreign exchange of Indian Rupee and stock market
price index in India. The rest of the paper has been arranged as follows: Section 2 gives the
methodological issues and Section 3 talks about the data and empirical results. Finally Section 4
concludes.
2. Methodological Issues
The dynamic linkage may simply be examined using the concept of Granger’s (1969, 1988) causality.
Formally, 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 being
identical. In other words, variable Xt fails to Granger-cause Yt if
where Pr(•) denotes conditional probability, ψ t is the set of all information available at time t and Ω t
is the information set obtained by excluding all information on Xt from ψ t. .
Testing causal relations between two stationary series Xt and Yt (in bivariate case) can be based on
the following two equations
p p
Y = α 0 + ∑ a k Y t − k + ∑ β k X t - k + ut (2)
t k =1
k =1
p p
Xt =ϕ0 + ∑ ϕ k Yt − k +
k =1
∑Φ
k =1
k X t − k + vt (3)
where p is a suitably chosen positive integer; a k’s and ßk’s, k = 0,1, … ..,p are constants; and ut and vt
usual disturbance terms with zero means and finite variances. The null hypothesis that Xt does not
2
Granger-cause Yt is not accepted if the ßk’s, k>0 in equation (2) are jointly significantly different
from zero using a standard joint test (e.g., an F test). Similarly, Yt Granger-causes Xt if the f k’s, k>0
coefficients in equation (3) are jointly different from zero.
It may be mentioned that the above test is applicable to stationary series. In reality, however,
underlying series may be non-stationary. In such cases, one has to transform the original series into
stationary series and causality tests would be performed based on transformed-stationary series. A
special class of non-stationary process is the I(1) process (i.e. the process possessing a unit root). An
I(1) process may be transformed to a stationary one by taking first order differencing. Thus, while
dealing with two I(1) process for causality, equations (2) and (3) must be expressed in terms of
differenced-series. However, if underlying I(1) processes are cointegrated, the specifications so
obtained must be modified by inserting the lagged-value of the cointegration relation (i.e. error-
correction term) as an additional explanatory variable (Engle and Granger, 1987). In other words,
equations (2) and (3) should be modified as
p p
∆Yt = α 0 + ∑
k =1
α k ∆ Yt− k + ∑
k =1
βk∆X t− k + δ ECT t -1 + ut (4)
p p
∆X t = φ + ∑ φ ∆Yt − k + ∑ Φ k ∆X t − k + η ECTt − 1 + v (5)
0 k =1 k k =1 t
where ∆ is the difference operator and ECTt-1 represents an error correction term derived from the
long-run cointegrating relationship between the I(1) processes Xt and Yt. This term can be estimated
by using the residual from a cointegrating regression. Clearly, if Xt and Yt are I(1) but not
cointegrated, the term ECTt-1 would be absent from equations (4) and (5).
The data used in this empirical study are daily stock market index and exchange rate (expressed in
Indian Rupee per U.S. dollar) for India. We use daily data for the period March 1993 to December
2002. We have chosen the stock index S&P CNX NIFTY of National Stock Exchange (NSE), an
index constructed with impact cost consideration, is one of the most robust indices available for
Indian stock market. We denote the chosen stock price index and exchange rate by NF and ER,
respectively. For the period from March 1993 to November 1995, we have used the simulated
NIFTY values calculated by NSE. From March 1993, India introduced unified exchange rate system
and started the process of exchange rate determination by market forces, though there has been
some sort of intervention by the central bank depending upon situations. However, as well known,
exchange rate decontrol has been a time consuming process and in 1994 India adopted the current
account convertibility. Keeping pace with the liberalization process in financial sector, exchange rate
liberalization has also been done in phases and hence it would be interesting to break the period into
various time buckets to see how the causal relationship has behaved over various time buckets.
Again, there has also been substantial regulatory changes taking place in capital market and over a
period of time, many international best practices have been implemented in a phased manner.
However, there have been concerns about the regulations on foreign investors’ownership of stocks
issued by domestic firms. Though it is a major issue that guides the flow of funds into various
markets, for India it has been a time consuming affair and even today we have sectoral limits for
foreign ownership.
The Granger causality test requires that all data series involved are stationary; otherwise the inference
from the F-statistic might be spurious because the test statistics will have nonstandard distributions.
Accordingly, we first employ the Augmented Dickey-Fuller (ADF) unit root tests (Dickey and Fuller,
1979, 1981) to check for stationarity for the exchange rate and stock market. Following the literature,
all data series have been transformed to natural logarithms. The equation used for conducting ADF
test has the following general structure
3
p
X t = α + β t + ρ X t− 1 + ∑ δ∆ X
k =1
k t− k + et (6)
In equation (6), if (i) β=0 and |ρ| <1, the series Xt is stationary; (ii) β=0 and ρ = 1 then the series is
an I(1) process; (iii) β≠ 0 and |ρ| < 1 then the series is trend-stationary (i.e. stationary around a
deterministic linear time trend). In Table 1, we present results of ADF tests for entire data period.
Estimated coefficients of time trend in equation (6) for different variables were statistically
insignificant and are not reported here. Empirical results indicate that there is a unit root in all the
logarithmic series (i.e. log(ER) and log(NF) series) and thus these log-level series can be described as
I(1) processes, indicating that their first-order differences (denoted by ∆Log(ER) and ∆Log(NF),
respectively) are stationary. Plot of ∆Log(ER) and ∆Log(NF) also appear to be oscillating around a
constant value (Figure 1). Further, application of ADF tests on ∆Log(ER) and ∆Log(NF) also
confirms the stationarity of underlying series (Table 1).
Log(ER) 5 -2.151
Log(NF) 1 -2.726
∆Log(ER) 5 -17.895*
∆Log(NF) 1 -45.545*
‘#’Optimal p is selected based on minimum Schwarz Information Criterion.
‘$’Critical Values of ADF test statistics for 1%, 5% and 10% level of significance are –
3.9676, -3.4144 and –3.1290, respectively.
‘*’Significant at 1% level.
4
Figure 1: Returns in Exchange Rate and Nifty Index
-2
-4
5
00 1
000 1
500 2
000
R
ET_
EXRA
TE
1
5
1
0
-5
-10
5
00 1
000 1
500 2
000
RE
T_NI
F T
Y
We now employ Johansen's (1991) maximum likelihood method to examine whether or not the
logarithms of exchange rate and stock price index for the country are cointegrated. The Table-2
reports the relevant results. As can be seen in the table, there is no cointegration vector between the
underlying series. Consequently, an error correction term need not be included in the Granger
causality test equations.
5
3.2 Linear Granger Causality Test Results
The test results of Granger-causality between exchange rates and stock prices are given in Table-3.
We experimented with a lag of 5-days from the consideration that there are 5 trading days in a week
and we hope that 5 days period would be adequate to get effects of one market to another under the
assumption of substantial informational efficiency.
For the full sample period, the reported F-values as well as p-values suggest that these two markets
did not have any causal relationship. If we go into individual years to see if the liberalization in both
the markets have brought them together or not, then also we do not see any significant causal
relationship between exchange rate and stock price movements except for the years 1993, 2001 and
2002, during when unidirectional causal influence from stock index return to return in forex market
is detected (as corresponding F-statistics are significant at 5 % level of significance). Very mild causal
influence in reverse direction is also found in some years (1997, 2002).
4. Conclusions
In this study, we examine the dynamic linkages between the foreign exchange and stock markets for
India. While the literature suggests the existence of significant interactions between the two markets,
6
our empirical results show that generally returns in these two markets are not inter-related, though in
recent years, the return in stock market had causal influence on return in exchange rate with
possibility of mild influence in reverse direction. These results have opened up some interesting
issues regarding the exchange rate and stock price causal relationship. In India, though stock market
investment does not constitute a very significant portion of total household savings compared to
other form of financial assets, it may have a significant impact on exchange rate movement as FII
investment has played a dominant role. The results, however, are tentative and there is a need to
undertake an in-depth research to address the issue.
References
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.
Ajayi, R. A. and M. Mougoue, 1996, “On the Dynamic Relation between Stock Prices and Exchange
Rates,”Journal of Financial Research 19, 193-207.
Bartov, E. and G. M. Bodnar, 1994, “Firm Valuation, Earnings Expectations, and the Exchange-Rate
Exposure Effect,” Journal of Finance 49, 1755-1785.
Bekaert, G. and C. R. Harvey, 2000, “Foreign Speculators and the Emerging Equity Markets,”
Journal of Finance 55, 565-613.
Bodnar, G. M. and W. M. Gentry, 1993, “Exchange Rate Exposure and Industry Characteristics:
Evidence from Canada, Japan, and the USA,” Journal of International Money and Finance 12, 29-45.
Chamberlain, S., J. S. Howe, and H. Popper, 1997, “The Exchange Rate Exposure of U. S. and
Japanese Banking Institutions,”Journal of Banking and Finance 21, 871-892.
Dickey, D. A. and W. A. Fuller (1979): Distribution for the estimators for auto-regressive time series
with unit root, Journal of American Statistical Association, Vol. 74, pp. 427-31.
Dickey, D. A. and W. A. Fuller (1981): Likelihood ratio statistics for auto-regressive time series with
unit root, Econometrica, Vol. 89, No. 4, July, pp. 1057-72.
Dornbusch, R. and S. Fischer, 1980, “Exchange Rates and Current Account,” American Economic
Review 70, 960-971.
Granger, C. W. J., 1969, “Investigating Causal Relations by Econometrics Models and Cross Spectral
Methods,” Econometrica 37, 424-438.
Griffin, J. M. and R. M. Stulz, 2001, “International Competition and Exchange Rate Shocks: A
Cross-Country Industry analysis of Stock Returns,”Review of Financial Studies 14, 215-241.
He, J. and L. K. Ng, 1998, “The Foreign Exchange Exposure of Japanese Multinational
Corporations,” Journal of Finance 53, 733-753.
Johansen, S., 1991, “Estimation and Hypothesis Testing of Cointegration Vectors in Gaussian Vector
Autoregressive Models,”Econometrica 58, 165-188.
7
Jorion, P., 1990, “The Exchange Rate Exposure of U.S. Multinationals,” Journal of Business 63, 331-
345.
Jorion, P., 1991, “The Pricing of Exchange Rate Risk in the Stock Market,” Journal of Financial and
Quantitative Analysis 26, 363-376.
Ma, C. K. and G. W. Kao, 1990, “On Exchange Rate Changes and Stock Price Reactions,” Journal of
Business Finance & Accounting 17, 441-449.
Pan, Fok & Lui (1999): Dynamic Linkages Between Exchange Rates and Stock Prices: Evidence
from Pacific Rim Countries, Shippensburg University Working Papers