A Study of the Causal Relationship between Real Exchange Rate of Renminbi and Hong Kong Stock Market Index

doi:10.4236/me.2012.35074

Paper Menu >>

Journal Menu >>

Modern Economy, 2012, 3, 563-566

http://dx.doi.org/10.4236/me.2012.35074 Published Online September 2012 (http://www.SciRP.org/journal/me)

A Study of the Causal Relationship between Real Exchange

Rate of Renminbi and Hong Kong Stock Market Index

Wai-Choi Lee1,2

1Department of Economics and Finance, Hang Seng Management College, Hong Kong, China

2School of Finance, Shanghai University of Finance and Economics, Shanghai, China

Email: [email protected]

Received May 23, 2012; revised June 20, 2012; accepted June 28, 2012

ABSTRACT

This paper attempts to study whether the real exchange rate of Renminbi and stock market index in Hong Kong are re-

lated to each other or not. The study uses cointegration and Granger causality tests on the monthly data of the real ex-

change rate of Renmnibi (RMB) in terms of Hong Kong dollars (REX) and Hang Seng Index (HSI) since the foreign

exchange reform of China in July 2005. The major findings of the study are a) the cointegration test shows that there

exists a long-run equilibrium relationship between REX and HSI; b) the error correction mechanism further evidences

that there is an error correction between REX and HSI in the short run; c) the Granger causality test indicates that there

is causal direction from RMB to HSI but not vice versa.

Keywords: Cointegration Test; Error Correction Mechanism; Granger Causality Test; Renminbi; Hang Seng Index

1. Introduction

In July 2005, the People’s Bank of China (PBoC) aban-

doned the 11-year-old peg of the Renminbi (RMB) against

the American dollar (USD) and replaced it with a man-

aged floating exchange rate based on market supply and

demand with reference to a basket of foreign currencies.

Not surprisingly, currencies in the basket included the

currencies of three trading partners—i.e., the USD, the

euro, and the Japanese yen. Under the new system, ad-

justments in the rate of exchange fluctuation will not only

govern the relationship between RMB and USD, but also

the relationship between RMB and other main currencies

such as the euro and the Japanese yen.

At the same time, the RMB was revalued by 2.1%—

from 8.28 RMB per USD to 8.11 per USD. In September

2005, the PBoC decided to adopt a narrow band of 0.3%

on either side of the central rate within which RMB can

fluctuate against USD. Since then, the RMB has gradu-

ally been revalued against USD. As of 22 May, 2012, the

exchange rate of RMB is raised to 6.33 per USD.

Since 1983, the monetary authority of Hong Kong has

maintained currency stability, defined as a stable external

exchange value of the currency of Hong Kong, in terms

of its exchange rate in the foreign exchange market against

the USD, at around HK$7.80 to US$1. As a result, the

exchange rates of Hong Kong dollar (HKD) in terms of

foreign currencies, such as RMB, follow the exchange

rates of USD in the same direction. As the USD depreciates

against RMB, the HKD depreciates against it accordingly.

The Hong Kong Stock Exchange is Asia’s third largest

stock exchange by market capitalization, behind the To-

kyo Stock Exchange and Shanghai Stock Exchange. Hang

Seng Index (HSI), the most widely quoted gauge of the

Hong Kong stock market, includes the largest and most

liquid stocks listed on Main Board of the Stock Exchange

of Hong Kong. It is a market capitalization weighted index

and comprises 48 constituent stocks of Hong Kong and

Mainland China enterprises. The aggregate market capi-

talization of these stocks accounts for over 65 percentage

of the total market capitalization in the Hong Kong Stock

Exchange.

Classical economic theory suggests a relationship be-

tween the stock market performance and the exchange

rate behaviour. Dornbusch and Fisher [1] and Aggarwal

[2] affirm that currency movements affect international

competitiveness and the balance of trade position, and

consequently the real output of the economy, which in

turn affects current and future cash flows of companies

and their stock prices.

Granger, Huang and Yang [3] find that currency de-

preciation could either raise or lower the value of Asian

companies during the Asian Crisis of 1997, depending on

whether the company mainly imports or mainly exports.

When the stock market index is considered, the net effect

cannot be predicted.

Yue, Cheng, Lee and Man [4] perform a panel data

analysis across four small export-oriented economies,

W.-C. LEE

564

namely, Hong Kong, Singapore, South Korea, and Taiwan.

The empirical results suggest that for small export-orie-

nted economies, a drop in the external of local currency

will result in an increase in exports and a rise in stock

market index.

Hong Kong is famous for being a small open economy,

allowing unrestricted flows of international goods and

funds. It is believed that the revaluation of the RMB

against the USD dampens the competitiveness of China’s

exports to the United States. Thus, the present and future

cash flows of Hong Kong and mainland export-oriented

enterprises are adversely affected. The stock market index

of Hong Kong is expected to decline in response to the

revaluation of the RMB against the USD/HKD. However,

the revaluation of the RMB against the HKD leads to

more capital flows from China to goods and assets markets

in Hong Kong. The stock market index of Hong Kong is

therefore expected to rise. The overall effect on Hong

Kong Stock Market Index is indeterminate without an

empirical study.

This paper attempts to examine whether the exchange

rate of RMB and HSI are related to each other since the

foreign exchange reform of China in July 2005 by using

cointegration and Granger causality tests. It is structured

as follows: Section 2 describes the data; the results of

cointegration test, error correction mechanism, and Granger

causality test are discussed in Sections 3, 4 and 5 respec-

tively; Section 6 concludes the paper.

2. Data Description

The monthly data of the nominal exchange rate of RMB

in terms of Hong Kong dollars (NEX), the composite

consumer price index of Hong Kong (CPIHK), the total

consumer price index of China (CPIChina), and Hang Seng

Index (HSI) are collected from Datastream for the time

period from August 2005 to March 2012. The real ex-

change rate of RMB in terms of HK dollars (REX) is

equal to NEX times CPIChina to CPIHK, i.e.,

China

CPI

XCPI

REX NE (1)

There are 80 observations to be used in this study. Ta-

ble 1 shows the descriptive statistics of REX and HSI.

Table 1. Descriptive Statistics of real exchange rate of RMB

in terms of HKD and Hang Seng Index.

Series REX HSI

Mean 1.155538 19947.90

Median 1.161332 20397.06

Maximum 1.243363 29540.78

Minimum 1.065253 12976.71

Standard deviation 0.052004 3604.71

Skewness –0.330679 0.097631

Kurtosis 2.113493 2.718397

Since the data used in this study are time-series data, if

they are non-stationary, regression will result in exagger-

ated results and spurious regression problem [5]. It is

essential to test the stationarity nature of the data before

the regression analysis.

REX and HSI are plotted in Figures 1 and 2. Both

time-series are not stationary because visually the mean,

variance, and autocovariances of the individual series do

not seem to be time-invariant. The Augumented Dickey-

Fuller (ADF) test for unit root [6] is used to test the sta-

tionarity of these time series.

Table 2 shows the results of ADF unit root test. The

ADF statistics of InHSI and InREX cannot reject the null

hypothesis of a unit root at 1% and 5% MacKinnon

critical value (–3.5188 and –2.9001 respectively). On the

other hand, the ADF statistics of InHSI and InREX

reject the null hypothesis at 1% and 5% significant levels

respectively. It is concluded that InHSI and InREX

are stationary and InHSI and InREX are of integrated

order of 1, in short, I(1).

3. Cointegration Test

According to the cointegration idea put forward by Engle

and Granger [7], although individual time-series variables

are non-stationary, there may still be a linear combina-

tion among the variables such that their stochastic trends

can be cancelled out. An Augmented Engle-Granger (AEG)

test is used to check for the stationarity of the term ut in

the following regression equation. The regression results

are shown in Table 3.

Real Exchange Rate of RMB in terms of HKD

0.95

1.05

1.1

1.15

1.2

1.25

1.3

8/1/05

2/1/06

8/1/06

2/1/07

8/1/07

2/1/08

8/1/08

2/1/09

8/1/09

2/1/10

8/1/10

2/1/11

8/1/11

2/1/12

Figure 1. Real exchange rate of RMB in terms of HKD,

August 2005-March 2012.

Hang Seng Index

5000

10000

15000

20000

25000

30000

35000

8/1/05

2/1/06

8/1/06

2/1/07

8/1/07

2/1/08

8/1/08

2/1/09

8/1/09

2/1/10

8/1/10

2/1/11

8/1/11

2/1/12

Figure 2. Hang Seng Index, August 2005-March 2012.

W.-C. LEE 565

Table 2. Results of ADF test.

Series ADF Test Statistic

InHSI –2.709003

InREX –1.064461

InHSI –3.709665

InREX –3.456199

Table 3. Regression results of Equation (2).

α β

Regression coefficients 9.528219 2.480117

Standard deviations 0.055285 0.367391

t-statistic 172.3472 6.750621

Probability (t-statistic) 0.0000 0.0000

Adjusted R-squared: 0.360691.

lnHSIt = α+ βlnREXt + ut (2)

Applying ADF unit root test on ut, the ADF test statis-

tic of the residual is –2.390807. The null hypothesis of a

unit root is rejected at 5% MacKinnon critical value

(–1.9443). It is concluded that HSI and REX are cointe-

grated. That is, there is a long-run equilibrium relation-

ship between HSI and REX. The regression on the two

variables in (2) is meaningful, not spurious; and there

exists a positive relationship between them as β, in Table

3, is positive.

4. Error Correction Mechanism

The error-correction mechanism (ECM) developed by

Engle and Granger [7] is used to test whether a disequi-

librium state will be adjusted to equilibrium in the short

run. The ECM in this study is as follows.

lnHSIt = α + β1lnREXt + β2ut – 1 +



t (3)

The regression results are shown in Table 4. The es-

timate of regression coefficient of ut – 1 is of correct sign

and found statistically significant at 1%. It is concluded

that there is error correction between REX and HSI in the

short run.

5. Granger Causality Test

The Granger Causality Test [8] is used to detect if there

is the cause-and- effect relationship between HSI and REX.

The test involves estimating the following regressions:

tit-i j

i=1 j=1

InHSI ΔInREX Δ



 



t-j 1t

InHSI u





j t-j2t

InREX +u

(4)

ti t-i

i=1 j=1

ΔInREX =ΔInHSI +Δ



 (5)

where it is assumed that the disturbances u1t and u2t are

uncorrelated.

The null hypotheses of the test and the regression results

by using 4 lags of each variable are shown in Table 5.

Table 4. Results of Equation (3).

α β1 β2

Regression coefficients–0.00403 –0.14138 –0.19262

Standard deviations0.00770 0.647329 0.05408

t-statistic –0.52314 –0.21841 –3.56184

Probability (t-statistic)0.60240 0.82770 0.00060

Adjusted R-squared: 0.125743.

Table 5. Results of Granger causality test.

Null Hypothesis F-Statistic Probability

(F-Statistic)

lnREXdoes not Granger

cause lnHSI 3.11554 0.02072

lnHSI does not Granger

cause lnREX 0.83532 0.50763

By using 4 lags, the null hypothesis of REX does not

Granger cause HSI is rejected as the probability value of

F-Statistic are close to 0. Another way of saying is that

REX Granger-causes HSI. However, HSI does not Gran-

ger-cause REX.

6. Conclusions

This paper attempts to examine whether the real exchange

rate of RMB and Hong Kong stock market index are re-

lated to each other. The result of cointegration test shows

that they are cointegrated. The regression result indicates

that there exists a positive relationship between them. In

addition, the result of error correction mechanism further

evidences that there is error correction between them in

the short run. The Granger causality test indicates that the

real exchange rate of RMB Granger-causes Hang Seng

Index.

To conclude, the movement of the real exchange rate

of RMB will lead to a movement of Hang Seng Index to

a certain extent. When RMB appreciates against HK dol-

lars, Hang Seng Index will rise. It might imply that the

positive effect of capital flows to goods and assets mar-

kets in Hong Kong on Hong Kong stock market is domi-

nant due to the revaluation of the RMB against the HKD.

On the other hand, Hang Seng Index does not Gran-

ger-cause the real exchange rate of RMB in terms of HK

dollars. It reflects the fact that since the HK dollar is

linked to the US dollar at a fixed rate, any economic and

financial changes in Hong Kong, such as a change in HSI,

does not have any impact on the exchange rates of other

foreign currencies in terms of US and HK dollars.

REFERENCES

[1] R. Dornbusch and S. Fisher, “Exchange Rates and the

Current Account,” American Economic Review, Vol. 70,

No. 5, 1980, pp. 644-547.

[2] R. Aggarwal, “Exchange Rates and Stock Prices: A Study

W.-C. LEE

566

of US Capital Markets under Floating Exchange Rates,”

Akron Business and Economic Review, Vol. 12, 1981, pp.

7-12.

[3] C. W. J. Granger, B. Huang and C. W. Yang, “A Bivari-

ate Causality between Stock Prices and Exchange Rates:

Evidence from Recent Asian Flu,” Quarterly Review of

Economics and Finance, Vol. 40, No. 3, 2000, pp. 337-

354. doi:10.1016/S1062-9769(00)00042-9

[4] H. Y. Yue, W. W. Cheng, W. C. Lee and K. K. Man,

“Interaction between Foreign Exchange Rate and Stock

Market Price Index in East Asian Economies: A Panel

Data Analysis,” 2012 Shanghai International Conference

on Social Science, Shanghai, 14-17 August 2012.

[5] C. W. J. Granger and P. Newbold, “Spurious Regression

in Econometrics,” Journal of Econometrics, Vol. 2, No. 2,

1974, pp. 111-120. doi:10.1016/0304-4076(74)90034-7

[6] D. A. Dickey and W. A. Fuller, “Distribution of the Esti-

mators for Autoregressive Time Series with a Unit Root,”

Journal of the American Statistical Association, Vol. 74,

No. 366, 1979, pp. 427-431.

[7] R. Engle and C. W. J. Granger, “Cointegration and Error

Correction: Representation, Estimate, and Testing,” Eco-

nometrica, Vol. 55, No. 2, 1987, pp. 251-276.

doi:10.2307/1913236

[8] C. W. J. Granger, “Investigating Causal Relations by

Econometric Models and Cross-Spectral Methods,” Eco-

nometrica, Vol. 37, No. 3, 1969, pp. 424-438.

doi:10.2307/1912791