Stock market and foreign exchange market integration in South Africa

Stock market and foreign exchange market integration in South Africa

World Development Perspectives 6 (2017) 32–34 Contents lists available at ScienceDirect World Development Perspectives journal homepage: www.elsevie...

219KB Sizes 0 Downloads 47 Views

World Development Perspectives 6 (2017) 32–34

Contents lists available at ScienceDirect

World Development Perspectives journal homepage: www.elsevier.com/locate/wdp

Case report

Stock market and foreign exchange market integration in South Africa Rajarshi Mitra Faculty of Economics, National Research University Higher School of Economics, 25/12 B. Pecherskaya Street, Nizhny Novgorod 603155, Russia

a r t i c l e

i n f o

Article history: Received 20 March 2017 Accepted 20 May 2017

JEL classification: F31 F41 O55 Keywords: Cointegration Exchange rate Stock index VECM

a b s t r a c t The total value of stock transactions in South Africa, in proportion to its GDP, increased from 2.64% in 1979 to 92.72% in 2014. The real effective exchange rate index decreased from 140.57 in 1979 to 77.62 in 2014. The results of empirical studies on the relation between exchange rate and stock transactions are mostly mixed and inconclusive. This paper applies the dynamic cointegration technique to time series data and re-examines the short-run and long-run association between the real effective exchange rate and the total value of stock transactions in South Africa over the post-Bretton Woods period 1979– 2014. Results indicate a significantly positive long-run relation between the real effective exchange rate and the total value of stock transactions in South Africa. Results also indicate lack of integration between stock transactions in the U.S. and South Africa. Ó 2017 Published by Elsevier Ltd.

1. Introduction The flow oriented theory of exchange rate behaviour postulates that there is a relationship between exchange rate movements and stock returns. Currency depreciation increases the demand for a nation’s exports and results in an improvement in the trade balance. Thus depreciation stimulates aggregate demand and increases the level of economic activity in a nation. Similarly, the stock oriented model and the arbitrage pricing theory hypothesize a relation between exchange rate movements and stock returns. Although the data appears to be consistent with the hypothesis of the flow oriented, stock oriented and arbitrage pricing models of exchange rate behaviour, the results of numerous regression based analyses have shown that the relation between exchange rate and stock returns vary with estimation technique, sample periods, data sources, and countries under study; thus, the effect of exchange rate movements on stock transactions can be either positive or negative. Even for the same country, the coefficient estimates and their significance have varied across different sample periods. Thus from a policy standpoint, the mixed and inconclusive results necessitate re-examination of the relation between exchange rate movements and stock transactions. The objectives of this paper are two-fold: one, examine the short-run and long-run effects of currency depreciation on the total value of stock transactions in South Africa, and

E-mail address: [email protected] http://dx.doi.org/10.1016/j.wdp.2017.05.001 2452-2929/Ó 2017 Published by Elsevier Ltd.

two, examine the degree of integration of the stock market in South Africa with the world capital markets, represented by the total value of stock transactions in the U.S. in this study.

2. Literature review The relation between exchange rate movements and stock returns depends on both economic and econometric factors and the results are mostly mixed. While Solnik (1987), Soenen and Hennigar (1988), Ma and Kao (1990) and Choi (1995) observed a negative relation between exchange rate and stock returns, Aggarwal (1981) and Roll (1992) reported a positive relation between the two variables. Phylaktis and Ravazzolo (2005) reported a positive relation between stock and foreign exchange markets when the U.S. stock market acts as a channel for the links. Chow, Lee, and Solt (1997) observed no relation between exchange rate and stock returns. Muhammad and Rasheed (2011) observed long-run relationship between exchange rate and stock prices for some South Asian countries. Olugbenga (2012) reported significantly positive relation between exchange rate and stock market in the short-run, and a significantly negative relation in the longrun. Some recent papers on stock market integration between countries include Chiang, Jeon, and Li (2007), Cheung, Cheung, and Ng (2007), Wang and Moore (2008), Cheung, Fung, and Tam (2008), Joshi (2011), Kim and Kim (2011) and Jeong (2012).

33

R. Mitra / World Development Perspectives 6 (2017) 32–34 Table 2 Johansen cointegration test.

3. Data and the model 3.1. Data The three variables in the model are the total value of stocks traded in South Africa (in proportion to GDP), the total value of stocks traded in the U.S. (in proportion to GDP), and the real effective exchange rate index. The total value of stocks traded in a nation is the number of domestic and foreign shares multiplied by their respective prices. The nominal effective exchange rate is the weighted average of the value of a nation’s currency against foreign currencies. The real effective exchange rate is the nominal effective exchange rate divided by a price deflator or an index of costs. The base year for the index is 2010. Annual data on the three variables are obtained from the World Development Indicators of the World Bank.

Maximum rank

Eigenvalue

Trace statistic

5% Critical value

0 1 2

– 0.54 0.21

42.81 15.39* 7.34

29.68 15.41 3.76

Table 3 Short-run coefficients. Coefficient ECMt1 Constant

DZ t ¼ A þ BZ t1 þ

l1 X bi DZ ti þ et

0.76 0.37

Standard error

Probability

0.17 1.85

0.00 0.84

Table 4 Long-run coefficients.

3.2. The model A VECM with l lags and r cointegrating vectors is estimated of the form:

*

STSA EX STUS Constant

Coefficient

Standard error

Probability

1.00 0.45* 0.01 10.01

– 0.17 0.17 –

– 0.01 0.93 –

ð1Þ

i¼1

3.3. Estimation method The VECM is estimated for short-run and long-run coefficients within the Johansen (1995) normalization framework. If r is the maximum rank of the cointegrating matrix, then a minimum r2 restrictions are to be imposed when estimating the long-run coefficients. The optimal number of lags is determined by both the Akaike Information Criterion (AIC) and the Schwartz Bayesian Information Criterion (SBIC). The main results are discussed in the ensuing sections. The diagnostic tests are lastly performed to examine autocorrelation at lag order, normality in error distribution and stability of the model. 4. The results The DF-GLS unit root test, proposed by Elliott, Rothenberg, and Stock (1996), is first performed. The minimum of Modified Akaike Information Criterion and Schwartz Criterion determines the optimal number of lags. The test examines the null hypothesis of a unit root in a variable against the alternative of trend stationarity. The results are reported in Table 1. The variables are I(1); thus the Johansen cointegration test is next performed to examine the existence of a long-run relationship between the variables. The results are reported in Table 2. The maximum rank of the cointegrating matrix is 1; thus, the model fails to reject the null hypothesis that there is one longrun relationship between the variables. Thus the VECM is to be estimated with 1 lag and 1 rank specification. Since the maximum rank is 1, then according to the Johansen (1995) normalization Table 1 Unit root test. DF-GLS STSA EX STUS First-Difference 4STSA 4EX 4STUS

Lags 2.34 2.48 2.35

1 1 1

5.27* 4.67* 3.99*

1 1 1

framework, a minimum 12 restrictions are to be imposed in order to determine the free parameters. The short-run and long-run coefficients are reported in Tables 3 and 4, respectively. Since the model is estimated at 1 lag, the short-run output contained only the adjustment coefficient and the constant term. The adjustment coefficient 0.76 in the cointegrating equation is negative and significant at 1% significance level. This ensures rapid adjustment toward long-run equilibrium, being corrected by 76% annually. The normalized cointegrating vector is (1, 0.45, 0.01, 10.01). The signs of the coefficients of the dependent variables will be reversed when interpreting the normalized cointegrating vector; thus the long-run effects of the total value of stock transactions in the U.S. (in proportion to GDP) and the real effective exchange rate on the total value of stock transactions in South Africa is given by

ST SA ¼ 10:01 þ 0:45 EX þ 0:01ST US

ð2Þ

The positive long-run coefficient 0.45 for the real effective exchange rate, significant at 1% significance level, indicates a positive relation between the real effective exchange rate index and the total value of stocks traded in South Africa. An increase in the index (currency appreciation) would lead to an increase in the total value of stock transactions in South Africa in proportion to its GDP. Thus there is a close association between the stock market and the foreign exchange market in South Africa. According to Phylaktis and Ravazzolo (2005), the relaxation of foreign exchange restrictions may be a necessary condition for the link between foreign exchange and stock markets. The degree of exchange rate flexibility may also have a significant role in the close association between the foreign exchange and stock markets in South Africa. The long-run coefficients indicate that stock transactions in the U.S. do not have any significant effect on the local stock market in South Africa. This indicates lack of financial integration of the South African stock market with the world capital markets, represented by the U.S. stock market in this study. This could be due to several reasons. As Phylaktis and Ravazzolo (2005) explained, apart from having access to capital market itself, access to local market information is necessary for international investments to take place there. According to Bekaert and Harvey (1995), Levine and Zervos (1996) and Bekaert and Harvey (2000), liberalization may

34

R. Mitra / World Development Perspectives 6 (2017) 32–34

Table 5 VECM Diagnostics. Autocorrelation

Chi-square statistic

Degrees of freedom

Probability

LM Lag 1

10.21

9

0.33

Normality

Chi-square statistic

Degrees of freedom

Probability

Jarque-Bera

1.93

2

0.38

R2: 0.37. VECM Unit Moduli for Eigenvalue Stability: 2.

not necessarily attract foreign investment due to home bias in equity portfolio, country-specific risks and lack of local market information on company stocks. Institutional rigidities and international differences in tax laws, as discussed in Feldstein and Horioka (1980), could impede capital mobility across countries and affect the degree of financial integration between countries. According to Levine and Zervos (1996), countries that publish information about a firm, such as its price-earnings ratio more comprehensively and internationally, will have better chances of developing a more globally integrated stock market than countries than do not make such information available to potential investors. The results of the LM test for autocorrelation, the Jarque-Bera test for normality in error distribution and stability of the model are reported in Table 5. The null hypothesis for the LM test is that there is no autocorrelation at lag order. The null hypothesis for the Jarque-Bera test is that the errors are normally distributed. As for the stability test, the VECM will be considered stable if a VECM with n endogenous variables and r cointegrating equations imposes no more than n  r unit moduli. The chi-square statistic 10.21 for the LM test is less than the 10% critical value 14.68 for 9 degrees of freedom. This confirms no autocorrelation at lag order. The chi-square statistic 1.93 is less than the 10% critical value 4.61 for 2 degrees of freedom; thus the errors are normally distributed. A VECM with three endogenous variables and one cointegrating relationship should not impose more than two unit moduli. The eigenvalue stability condition is satisfied. 5. Conclusion In light of mixed and inconclusive results in the existing literature, this paper has re-examined the relation between the real effective exchange rate and the total value of stock transactions in South Africa over the post-Bretton Woods period 1979–2014. The cointegration technique is applied since it addresses problems with non-stationarity in time series data. The results indicate a close association between the stock and foreign exchange markets in South Africa. Currency appreciation is found to significantly increase the total value of stock transactions in South Africa. The positive relation could arise due to increases in stock returns and economic activity. Results also indicate lack of a significant relation between stock transactions in the U.S. and South Africa. From a policy perspective, the significantly positive relation between the stock market and the foreign exchange market in South Africa indicates that exchange rate as a tool to increase

foreign investment in the country. The lack of integration between the stock markets in the U.S. and South Africa indicates that appropriate monetary and fiscal policies must be designed and implemented in a way that will integrate the South African market with the international capital market. The model may be modified and extended by including additional control variables, such as institutional strength, allowing for structural breaks and examining the line of causality between the variables for robust regression analyses. References Aggarwal, R. (1981). Exchange rates and stock prices: A study of the US capital markets under floating exchange rates. Akron Business and Economic Review, 7–12. Bekaert, G., & Harvey, C. R. (1995). Time-varying world market integration. Journal of Finance, 50, 403–444. Bekaert, G., & Harvey, C. R. (2000). Foreign speculation and emerging equity markets. Journal of Finance, 55, 565–613. Cheung, Y. L., Cheung, Y. W., & Ng, C. (2007). East Asian equity markets, financial crises, and the Japanese currency. Journal of Japanese and International Economies, 21, 138–152. Cheung, L., Fung, L. and Tam, C.S. (2008). Measuring financial market interdependence and assessing possible contagion risk in the EMEAP region. Hong Kong Monetary Authority Working Paper No. 18/2008. Chiang, T. C., Jeon, B. N., & Li, H. (2007). Dynamic correlation analysis of financial contagion: Evidence from Asian markets. Journal of International Money and Finance, 26, 1206–1228. Choi, J. J. (1995). The Japanese and US stock prices: A comparative fundamental analysis. Japan and the World Economy, 7, 347–360. Chow, E. H., Lee, W. Y., & Solt, M. S. (1997). The exchange rate risk exposure of asset returns. Journal of Business, 70, 105–123. Elliott, G., Rothenberg, T. J., & Stock, J. H. (1996). Efficient tests for an autoregressive unit root. Econometrica, 64, 813–836. Feldstein, M., & Horioka, C. (1980). Domestic saving and international capital flows. Economic Journal, 90, 314–329. Jeong, J. (2012). Dynamic Stock Market Integration and Financial Crisis: The Case of China, Japan, and Korea. In Working Paper, Available at CFRN: http://www. cfrn.com.cn/getPaper.do?id=3370. Johansen, S. (1995). Likelihood-Based Inference in Cointegrated Vector Autoregressive Models. Oxford: Oxford University Press. Joshi, P. (2011). Return and Volatility Spillovers Among Asian Stock Markets Available at http://sgo.sagepub.com/content/early/2011/06/10/ 2158244011413474. Kim, B., & Kim, H. (2011). Spillover Effects of the US Financial Crisis on Financial Markets in Emerging Asian Countries. Auburn Economics Working Paper Series 2011–04, Department of Economics, Auburn University. Levine, R., & Zervos, S. (1996). Capital control liberalisation and stock market development. The World Bank group Policy Research Working Paper no.1622. Ma, C. K., & Kao, G. W. (1990). On exchange rate changes and stock price reactions. Journal of Business Finance and Accounting, 17, 441–449. Muhammad, N., & Rasheed, A. (2011). Stock prices and exchange rates: Are they related? Evidence from South Asian Countries. Pakistan Development Review, 41, 535–550. Olugbenga, A. A. (2012). Exchange rate volatility and stock market behaviour: The Nigerian experience. Research Journal of Finance and Accounting, 3, 88–96. Phylaktis, K., & Ravazzolo, F. (2005). Stock prices and exchange rate dynamics. Journal of International Money and Finance, 24, 1031–1053. Roll, R. (1992). Industrial structure and the comparative behaviour of international stock market indices. Journal of Finance, 47, 3–41. Soenen, L. A., & Hennigar, E. S. (1988). An analysis of exchange rates and stock prices: The US experience between 1980 and 1986. Akron Business and Economic Review, 71–76. Solnik, B. (1987). Using financial prices to test exchange rate models: A note. Journal of Finance, 42, 141–149. Wang, P., & Moore, T. (2008). Stock market integration for the transition economic: Time-varying conditional correlation approach. The Manchester School, 76, 116–133.