The transmission of market shocks and bilateral linkages: Evidence from emerging economies

The transmission of market shocks and bilateral linkages: Evidence from emerging economies

    The Transmission of Market Shocks and Bilateral Linkages: Evidence from Emerging Economies Faruk Balli, Hatice O. Balli, Rosmy Jean L...

588KB Sizes 0 Downloads 14 Views

    The Transmission of Market Shocks and Bilateral Linkages: Evidence from Emerging Economies Faruk Balli, Hatice O. Balli, Rosmy Jean Louis, Tuan Kiet Vo PII: DOI: Reference:

S1057-5219(15)00143-X doi: 10.1016/j.irfa.2015.08.010 FINANA 891

To appear in:

International Review of Financial Analysis

Received date: Accepted date:

3 August 2015 11 August 2015

Please cite this article as: Balli, F., Balli, H.O., Louis, R.J. & Vo, T.K., The Transmission of Market Shocks and Bilateral Linkages: Evidence from Emerging Economies, International Review of Financial Analysis (2015), doi: 10.1016/j.irfa.2015.08.010

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT The Transmission of Market Shocks and Bilateral Linkages: Evidence from Emerging Economies Hatice O Balli2

Rosmy Jean Louis3

Tuan Kiet Vo4

RI P

T

Faruk Balli1

SC

A BST R AC T

NU

The linkage between emerging and developed economies spans beyond the usual trade in goods and services. Underlying trade is the flow of capital for foreign direct investment and for speculation in markets, which renders emerging economies vulnerable to shocks from the developed world. As such,

MA

equity return volatility in emerging markets is partly attributable to this dependence. To gauge the importance of bilateral economic and cultural factors in driving economic integration, we adopt a twostep process. First, we use Diebold and Yilmaz’s spillover index methodology to extract spillover indices

ED

representative of the return volatility spillover effects of the United States, the developed portion of the Euro area, and Japan on financial markets in Asia, the Gulf Cooperation Council countries, Eastern and Central Europe, Africa, and Latin America. Second, we test whether these indices are governed by

PT

economic and cultural factors. Our results show that the spillover effects vary across markets and that a strong correlation exists with the volume of trade, security investment, common language, distance, and

CE

market capitalization.

AC

JEL Classification: F21, F36, G12, G15 Keywords: bilateral trade, equity market, GARCH, market integration, security investment, spillover, volatility

1

School of economics and finance, Massey University, Auckland. E mail: [email protected]

2

School of economics and finance, Massey University, Auckland. Department of Economics and finance, Vancouver Island University, Canada.

3

4

School of economics and finance, Massey University.

1

ACCEPTED MANUSCRIPT 1.

I NTRODU CTION

Advances in technology of information, mainly the internet and, more recently, byproducts such as YouTube, Facebook, Twitter, etc., have contributed a great deal to a

T

more connected world. The Information Age has without a doubt complemented policy

RI P

initiatives towards market liberalization over the last four decades, giving rise to the wave of the globalization of national capital markets. In general, the markets started to

SC

play an active role in this movement in the early 1980s with the advent of policies toward interest and exchange rate deregulation, as well as efforts towards reducing or

NU

removing barriers to foreign investments (Bekaert and Harvey, 1995, 2000). These efforts subsequently led to a spectacular development of international exchanges across

MA

countries, particularly in the developed world (Bekaert et al., 2002). As reported by the International Finance Corporation in their 2008 annual report, net flows of private capital towards emerging markets reached USD$616 billion in 2007. In recent years

ED

emerging markets have accounted for about 50% of the world's economic growth. Along with the movement towards the globalization of domestic markets, worldwide

PT

economic areas have continued to develop their institutional aspects, as shown by the introductions of a number of regional economic agreements (European Union, ASEAN,

CE

(GCC), etc.). These regional trade agreements carry the seeds of greater openness, which could translate into more competitiveness at the world level when member countries jointly exert effort and synergy. Several emerging regions such as Central and Eastern

AC

Europe, Asia, and South America are also in keeping with these dynamics, both at the regional and the global scales. However, the relationship between global and regional integration is not consistent across different areas, nor is the speed of the financial integration process and procedures consistent over time. In some regions such as the GCC and Asia, international integration is ahead of the regional integration, whereas in areas such as Eastern and Central Europe and Latin America, it is the opposite that is observed. The benefits of globalization and financial integration, in particular, are well documented in the literature: higher potential for risk sharing, a more efficient allocation of capital, and brighter economic growth prospects for emerging markets. However, a number of unwanted side effects come along, including higher financial vulnerability in the event of an economic crisis, and trade disparities with developed countries ( Levine and Zervos, 1998; Stiglitz, 2002) As the emerging markets mature

2

ACCEPTED MANUSCRIPT and co-move with the world markets, they become more responsive to the volatility of equity markets elsewhere. This integration with world markets has the potential to constrain international portfolio diversification (Neaime, 2002, Balli et al. 2014b). To

T

what extent these may occur depends entirely on the level of financial integration. This

RI P

is a void that a detailed assessment can fill for these markets. Such studies can shed light on fundamental perspectives such as the determinants and effects of current trends in

SC

financial integration on risk premia and cost of capital.

Research into cross-border linkages in emerging equity markets has been motivated by growth and increasing openness, as well as by the speed and severity with which

NU

past financial crises in these markets have spread to other countries. Over the last two decades or so, a variety of papers have provided a general understanding on the

MA

integration of emerging markets. Bekaert and Harvey (1995, 1997, 2000), Bekaert et al. (2005) and Carrieri et al. (2007) studied the implication of increasing integration with global markets on local returns, volatility, and cross-country correlations, covering

ED

a wide range of emerging markets comprising Asia, Eastern and Central Europe, Latin America, and the Mediterranean area.

Other studies of emerging stock markets,

PT

however, specifically looked at specific regions. Scheicher (2001), and Yang et al. (2006), Balli and Balli(2011) and Balli et al.(2013b) studied the extent and effects of

CE

stock market integration in the Europe, both regionally and globally. Chen et al. (2002), Abugri (2008), Susmel (2001), and Diamandis (2009) examined the

AC

evidence of regional linkages among Latin American equity markets. Neaime (2002, 2006, 2012) and Floros (2008), Balli et al.(2013a) extensively concentrated on the booming stock markets in the Middle East, while Ng (2000), Tay and Zhu (2000), Worthington and Higgs (2004), Caporale et al. (2005; 2006), Engle et al. (2008), Yilmaz (2010) and Balli et al.(2014) focussed on the dynamics of stock markets in emerging Asia. However, little attention has been paid to the cross-regional dynamics of emerging markets’ integration with the world economy, which, according to Bracker et al. (1999) can open a whole new strand of studies on how macroeconomic, social, and cultural factors affect these markets, particularly their bond and stock markets. Few studies thus far have attempted to gauge the extent of these countries’ stock market integration that is attributable to macroeconomic factors. Chen and Zhang (1997)

3

ACCEPTED MANUSCRIPT explained the links between stock market volatility and the intensity of bilateral trade. They correlated the emerging stock markets of South Korea, Taiwan, Thailand, Malaysia, Philippines, and Mexico with the developed markets of the United States (US), Canada,

T

Japan, Hong Kong, Singapore, Australia, New Zealand, Austria, the United Kingdom, and

RI P

some developed European countries. They showed that stock market interdependence is positively correlated with the magnitude of trade. In fact, trade explains 5–40% of the variation in the correlation, depending on the measure of correlation used. In a similar

SC

study, Bracker et al. (1999) used daily returns to construct a time series of the correlation matrix and found that the matrix had changed substantially over time. In

NU

fact, the degree of interdependence was positively correlated with market volatility and trend, but negatively correlated with exchange rate volatility, real interest rate

MA

differentials, the return on the world index, and the term structure differentials. Although the literature has pointed to stock markets being interdependent and driven by economic factors (Bekaert and Harvey, 1997) one issue that remains

ED

unresolved is whether the factors that drive co-movement in more mature markets are also common to emerging markets). Contributions by Pretorius (2002) and, more

PT

recently, Lucey and Zhang (2010) attempts in this direction. In this vein, our paper complements the existing literature in conducting a two-step analysis. First, we quantify

CE

the spillovers from major countries/regions (US, the Euro area, and Japan) on emerging markets using the methodology proposed by Diebold and Yilmaz (2009) and extract a

AC

variety of shocks affecting emerging markets. Second, to capture the links between emerging and developed markets, we use a cross-sectional regression technique to test the relevance of macroeconomic factors, cultural affinities, and geographic distance in explaining the shocks. The results show that bilateral trade, security investments, a common language, and market capitalization are important determinants of shock spillovers to emerging markets. We also find evidence that geographical distance and, to a lesser extent, colonial ties matter. This, to our knowledge from our reading of the existing literature, is quite a novel finding.

2.

D ATA

AND

D ESCRIPTIVE S TATIS TICS

4

ACCEPTED MANUSCRIPT The data for this study come from a number of sources. The list of emerging markets (39 in total) is based on the 2012 International Monetary Fund (IMF) listing and the Financial Times Stock Exchange classification of markets. Of these markets,

T

seven are classed as the largest emerging and developing economies by either nominal

RI P

or inflation-adjusted gross domestic product. The sample of 39 markets was subsequently classified into five groups based on their geographic regions. This grouping enables us to distinguish among differential volatility effects that may emerge

SC

due to differences in locations and economic areas. Accordingly, our study focuses on Asia, the GCC, Eastern and Central Europe, Africa, and Latin America.

NU

The dataset includes bilateral trade volumes between emerging markets and the developed economies of the US, Europe (the Euro area), and Japan, as well as weekly

MA

equity return data. In addition, we have gathered data on geographic distances as well as data on the equity investments and debt securities issued by emerging markets being

ED

held by investors in developed economies.

Quarterly bilateral trade data came from the IMF’s Direction of Trade Statistics

PT

database, equity data in US dollars was taken from the Morgan Stanley Capital International database. Annual aggregate values of equity and debt security investments

CE

were also extracted from the IMF’s Coordinated Portfolio Investment Survey database, and the geographic distance data were collected from the French Research Center in International Economics(CEPII). The data stretch over the period 1990–2013, which

AC

encompasses the recent financial crisis. Table 1 presents the descriptive statistics for the weekly equity returns. It shows that the mean values vary between -0.70% (Ukraine) and 0.24% (Mexico and Oman). The variation in returns as measured by the standard deviation ranges from 2.42% (Tunisia) to 7.54% (Venezuela). Higher degrees of volatility are observed for countries with less stable economic conditions in most cases. For these countries, stock prices are more vulnerable to unusual economic disturbances. The statistical distributions show that most of the returns are skewed to the left and all suffer from excess kurtosis, which is quite high in some cases. The last four columns of Tables 1 and 2 report the Ljung and Box (1978) Portmanteau Q and Q† (for the squared data) test statistics for first- and second-moment dependencies in the distribution of the emerging market returns. In many cases, the Q and Q† statistics are significant, suggesting that the equity returns are serially correlated and subject to

5

ACCEPTED MANUSCRIPT strong second-moment dependencies. Table 2 also contains the summary statistics of bilateral trade, investments, and geographic distance between the markets on a crosssection basis.

T

For each of the 39 selected emerging markets, we compute their total trade

RI P

(exports plus imports) with each developed market as a share of their overall trade with the rest of the world on a yearly basis. We follow a similar procedure for the equity and

SC

debt investment measures by looking at the relative importance of each emerging market in the investment portfolio of developed markets. We find that, on average, the bilateral trade measures range from 0.71% (imports from Japan) to 24.77% (imports

NU

from the Euro area). This is a huge gap in terms of trade ties. One likely explanation of this difference may have to do with geographical distance and colonial ties. Across

MA

countries, we observe the relative importance of trade to be even more dispersed, ranging from 0.05% (almost no trade) to approximately 80%. These statistics are quite similar for the debt and equity investment measures: the mean values range from

ED

5.43% to 32.53% for debt investments and from 0.79% to 29.31% for equity

M ETH ODOLOGY

CE

3.

PT

investments.

3.1. THE SPILLOVER INDEX

AC

As a first step towards gauging the spillover effects, we use Diebold and Yilmaz’s (2009) methodology to extract spillover indices representative of the return volatility spillover effects of the US, the Euro area, and Japan on the financial markets of Asia, the GCC, Eastern and Central Europe, Africa, and Latin America. Basically, the stock market returns for all countries are represented as an N-variable VAR. For each stock market , we add the share of its forecast error variances coming from shocks originating from stock market j for all  ≠ . Next, we add across all  = 1, … ,  in order to obtain a single spillover index. Quantitatively, the spillover index is the sum of all non-diagonal elements of the forecast error variance–covariance matrix. For simplicity of exposition, we use a covariance stationary first-order bivariate VAR given by: ௧ = ௧ିଵ + ௧,

(1)

6

ACCEPTED MANUSCRIPT where ୲ = ଵ௧ , ଶ୲ , Φ is a 2 × 2 parameter matrix, and the vector of error terms ε୲ has zero mean. ୲ is either a vector of stock returns or volatilities. On the assumption that the VAR has stationary covariance, its moving average exists and is given by:

T

௧ =  ௧ ,

(2)

RI P

where  = ( − Φ )ିଵ . Using the Cholesky decomposition of the covariance matrix of ε୲ , the moving average can be rewritten as:

SC

௧ =  ௧ ,

(3)

where  = ( )୲ିଵ, ୲ = ୲ ε୲, ୲ ୲ᇱ = , and ୲ିଵ is the lower-triangular

NU

Cholesky factor of the covariance matrix of ε୲ . As a result, ୲ represents the orthogonalized structural shocks, with zero mean and a matrix of variance–covariance

MA

with ones as diagonal elements and zeroes elsewhere. For the one-step-ahead forecast, the optimal forecast5 is given by:

ED

௧ାଵ,௧ =  ௧

(4)

with the corresponding one-step-ahead error vector

PT

଴,ଵଵ ௧ାଵ,௧ = ௧ାଵ − ௧ାଵ,௧ = ଴ ௧ାଵ =  ଴,ଶଵ

଴,ଵଶ ଵ,௧ାଵ ଴,ଶଶ  ଶ,௧ାଵ ,

CE

ᇱ  = ଴ ᇱ଴. with the covariance matrix ௧ାଵ ௧ାଵ,௧

(5) (6)

ଶ ଶ + ଴,ଵଶ , Therefore, the variance of the one-step-ahead error forecast of ଵ୲ is ଴,ଵଵ

AC

ଶ ଶ and that of ଶ୲ is ଴,ଶଵ + ଴,ଶଶ .

Intuitively, we are interested in finding what fraction of the one-step-ahead error variance in forecasting ଵ is due to own shocks (ଵ ) or spillover shocks from ଶ . Likewise, what fraction of the one-step-ahead error variance in forecasting ଶ is due to own shocks (ଶ ) or to spillover shocks from ଵ ? ଶ ଶ In the bivariate case, the aggregate spillover is ଴,ଵଶ + ଴,ଶଵ , whereas the total

forecast error variation is given by   ଴ ᇱ଴ = aଶ଴,ଵଵ + aଶ଴,ଵଶ + aଶ଴,ଶଵ + aଶ଴,ଶଶ = . Hence, the spillover index ratio is:

 =

మ   మ బ,భమ బ,మభ

  బ ᇲబ

× 100.

(7)

5 We follow Diebold and Yilmaz (2009) and have employed the Wiener–Kolmogorov linear leastsquares forecasting methodology.

7

ACCEPTED MANUSCRIPT By generalizing this process for a one-step-ahead forecast with a ୲୦ -order Nvariable VAR, the spillover index can be represented as: మ ∑ಿ ೔,ೕసభ బ,೔ೕ

× 100.

(8)

T

೔ಯೕ

  బ ᇲబ

RI P

 =

3.2. CROSS-SECTION ANALYSIS

SC

Upon determining the magnitude of shocks using the variance ratio analysis, we carried out a second layer of analysis in investigating the possible underlying we conjecture that the impacts of shocks on

NU

determinants of spillovers. Hence,

emerging markets are linked to the volume of bilateral trade (TRADE), the amount of

MA

security investments (INV), a common language (LANG), market capitalization (CAPT), historical and colonial dependence to date (COL), as well as the time-invariant distance measure (DIST). The cross-sectional regression equation for the emerging markets is

ED

given by:

 ௝ = ଴ + ଵ  ! ௝ + ଶ " ௝ + ଷ # ௝ + ସ $ %௜ + ହ $ ௝ + ଺ "& ௝ + ௜  ௜ ௜ ௜ ௜ ௜ ௜

PT

(10)

 ௜ is the calculated variance ratio indicative of the relative magnitude of shocks where 

CE



from developed market  to emerging market  and  !௜ is the magnitude of ௝

AC

international trade between emerging country  and developed country . We use the volume of exports as a proxy since there is evidence of a high correlation between either exports or imports and the magnitude of trade. Since our focus is on spillover effects, the inflow of export revenues is a more suitable measure to capture trading activities. "௜ is a measure of emerging market securities held by developed country  ௝

as investment. #௜ is a dummy variable that takes 1 if a common language exists ௝

between emerging market i and developed country  or 0 otherwise. $ %௜ is the logarithm of the volume of market capitalization representing the relative size of the emerging equity market . $ ௜ is a dummy variable capturing colonial ties. This ௝

variable takes 1 when the characteristic is present or zero otherwise.

"&௜ is the ௝

time-invariant distance in kilometres between the capital cities of countries  and j.

8

ACCEPTED MANUSCRIPT The cross-sectional estimations of Eq. (10) are carried out in two steps. First, we document the determinants of the spillover effects of each developed market on all emerging markets based on a sample of 39 observations. Second, the data for all three

T

major economies’ influence on emerging markets are pooled together to test the

RI P

relative importance of the underlying determinants. This estimation is based on a pool of 117 (= 39 × 3) observations and therefore encompasses the overall influence of the

E MPIRICAL A NALYSI S

NU

4.

SC

three developed markets’ spillovers as a group on the emerging markets.

4.1. SPILLOVER MODEL

MA

We estimate spillovers over the sample period and present the empirical results in Table 3, which is organized by region.6 We find a higher magnitude of spillovers from the US for all Asian markets (more than 20% on average), and the Filipino and

ED

Bangladeshi markets being affected the most by the US markets. This result is partly consistent with studies by Bekaert and Harvey (1997) and Masih and Masih (1999),

PT

which documented a high level of interdependence and co-movement between the equity markets of the Philippines and the developed markets, particularly the US

CE

market. We also observe that Euro are markets have, on average, less than a 10% spillover effect on Asia markets, whereas the Japanese effect is around 15%. More

AC

importantly, the spillover effect of the US market is higher for countries that have greater bilateral trade and investment shares with the US (according to Figures 1a and 2a) . Figure 2a shows that Bangladesh and the Philippines have the highest bilateral trade share with the US (more than 30% of these countries’ trade). This is also reflected in spillover tables where these countries’ markets have been affected by US markets the most among Asian markets. For the GCC region, we have similar results for the Asian markets.

There are

significant spillover effects from the developed markets, except to Bahrain. The high degree of spillovers to the GCC is consistent with the findings of Neaime (2006), Yu and Hassan (2008), and Khalifa et al. (2012) on the global integration of the GCC– Middle East and North Africa region. The magnitude of the spillovers from the Euro 6

We are grateful to Diebold and Yilmaz (2009) for making their codes available to the public. These codes were modified to suit our purpose.

9

ACCEPTED MANUSCRIPT area and the US are dominate the volatility GCC markets, compared to Japan’s effect. Consistent with our hypothesis, the trade and investment shares of the US and Euro are with GCC economies (see Figure 1b and 2b), might be considered an important reason

T

for the higher magnitude of the spillovers from these countries to GCC markets. To

RI P

explore this relationship further, Figure 2b shows that Qatar, Kuwait, and the United Arab Emirates (UAE) have greater trade shares with Japan compared to the other GCC counties. More interestingly, spillovers from Japan to these markets are much higher

SC

compared to the other countries, which again might be considered as an evidence to

NU

relate bilateral links(trade, in this case) with spillover magnitudes.

For Eastern and Central European markets, greater spillovers magnitudes from the

MA

Euro area and US markets are detected, whereas the Japanese effect on these markets is relatively small. The bilateral trade and investment links between Eastern and Central

ED

Europe and the US, Euro area and Japanese markets are presented in Figures 1c and 2c. Both figures show that the US and Euro are markets have greater shares in the Eastern

PT

and Central European trade investment volumes-compared to Japan, which is consistent with the spillover magnitudes presented in Table 3.

CE

For the Latin American countries, spillovers from US markets are around 20% in average, whereas the spillovers from Euro area are less than 10%; those from Japan are

AC

much lower. Indeed, the bilateral trade and investment links between the US and Latin American countries are in greater in magnitude compared to the Euro area and Japanese markets (Figures 1e and 2e), which is consistent with the magnitudes of the spillovers in Table 3. Table 3a contains the estimations of Eq(8)

with a sub-period. We restrict the

estimations to post Global financial crises period (2008-2013) and presented the result accordingly. At the first glance, the decomposition of the shocks has not been changed much from Table 3. However, we observe that for almost all markets the extent of the effect of US and Euro markets increased substantially. This finding indicates the contagion from developed to emerging markets, and the effect is higher from US and Euro markets.

10

ACCEPTED MANUSCRIPT Overall, we observe that the US market has the dominant spillover effect on almost all markets around the world, compared to the Euro area and Japanese markets. Japan’s market spillovers affect East Asian and GCC markets, whereas the Euro are market has

T

relatively more effect in Eastern European and African markets. More importantly, the

RI P

spillover estimates suggest a relationship between the magnitudes of spillovers (Table

SC

3) and bilateral trade and investment linkages (Figures 1 and 2).

It seems the there is a strong relationship between bilateral trade and investment

NU

linkages and spillovers. Taking this relation into account, we investigate the determinants of the magnitudes of the spillovers from developed countries to the

4.2. CROSS-SECTION ANALYSIS

MA

emerging markets.

ED

In Table 4, we present estimates of the underlying determinants of spillovers from the US, the Euro area, and Japan to the 39 emerging markets, as per the cross-section

PT

model of Eq. (10). As described in the methodology above, we hypothesize that trade, investment, a common language, capitalization, colonial ties, and distance are the most

CE

likely determinants of the variance spillover ratios or shocks. Indeed, we find that trade is statistically significant in explaining spillovers from

AC

each of the major economies at the 1% level. A 1% increase in trade produces positive spillovers of 0.19% from the US, 0.41% from the developed Euro area, and 0.64% from Japan. It is quite intriguing that the Euro area and Japan are, respectively, two and three times more likely to spread shocks to emerging markets via trade than the US. This finding is consistent with Bracker et al. (1999) and earlier works by Bodurtha et al. (1989) and Campbell and Hamao (1992), which document a link between equity market integration and bilateral trade among other macroeconomic determinants. The coefficients of investment are statistically significant for the US and the Euro are segment spillovers but not for Japan. The spillover impacts are, respectively, 0.34% and 0.21% for the US and the Euro area due to a 1% increase in investment. These results are consistent with Jinjarak et al.'s (2011) account of an association between international investment funds and local market returns and Umutlu et al.'s (2010) study, which shed light on the effects of market liberalization on stock return volatility

11

ACCEPTED MANUSCRIPT in emerging markets. A common language is only significant for spillovers from the US economy at the 10% level, which, by and large, tends to suggest that market dynamics have little to do with the ability to communicate in the same language.

Market

T

capitalization as a proxy for measuring the financial depth of the emerging market has a

RI P

negative but statistically significant coefficient estimate at the 10% level for the US and the Euro area, thereby suggesting that as the emerging markets get stronger, they become less vulnerable to shocks from these two economies. Last but not least, we find

SC

that distance is statistically significant for the Euro area and Japan at the 10% and 5% levels, respectively. In all three cases, however, the coefficient estimate is negative,

NU

suggesting that the further away countries are from each other, the lower the resulting shock spillovers.

MA

To test the robustness of our results and to determine whether there is a pattern common to all three developed economies, we stack the variables and re-estimate Eq. (10). In other words, we investigate the underlying determinants of spillovers from the

ED

developed markets as a group to the emerging markets. Whether we test the significance of the variables individually or jointly, the results are robust. As shown in

PT

Table 5, trade, investment securities, market capitalization, and language are statistically significant at either the 1% or 5% level. Colonial ties and distance are

CE

statistically significant at the 10% level when estimated jointly with other variables but, when tested individually, only distance is significant at the 5% level. All variables have

AC

the appropriate expected signs. The only minor discrepancy from the initial result is that common language is now significant at the 1% level. This result therefore supports the findings by Bodurtha et al. (1989), Campbell and Hamao (1992), Bracker et al. (1999), and Goetzmann et al. (2005) that common language is important in explaining the co-movement of international stock markets and

long-term global

market correlations.

5.

C ONCLUSIONS In this paper, we aimed to investigate the dynamic links between equity market

integration in five emerging market regions and developed markets. Building upon the methodology proposed by Diebold and Yilmaz (2009), we carried out two layers of

12

ACCEPTED MANUSCRIPT analysis. First, we investigated the influence of each developed market on the full set of emerging markets to document the actual determinants of the variance spillovers. Second, in search of robustness, we pooled the overall data to uncover the existence (or

T

lack thereof) of a general pattern in spillovers. In short, we found that trade, investment,

RI P

market capitalization, common language, colonial ties, and distance are all important, to differing degrees, in explaining shock spillovers from developed to emerging markets. Therefore, we conjecture that this paper has made a distinctive contribution to the

SC

literature on equity market integration by successfully documenting that shock spillovers are governed not only by bilateral economic factors such as bilateral trade

NU

and investment securities but also by cultural factors such as a common language and geographical proximity as captured by distance.

MA

R EFERENCES

Abugri, B. A. (2008). Empirical relationship between macroeconomic volatility and stock returns:

ED

evidence from Latin American markets. International Review of Financial Analysis, 17(2), 396– 410.

PT

Ayuso, J., Blanco, R. (2001). Has financial market integration increased during the nineties? Journal of International Financial Markets, Institutions and Money, 11(3), 265–287. Baca, S. P., Garbe, B. L., Weiss, R. A. (2000). The rise of sector effects in major equity markets. Financial

CE

Analysts Journal, 34–40.

Balli, F., & Balli, H. O. (2011). Sectoral equity returns in the Euro region: Is there any room for reducing

AC

portfolio risk?. Journal of Economics and Business, 63(2), 89-106. Balli, F., Basher, S. A., & Louis, R. J. (2013a). Sectoral equity returns and portfolio diversification opportunities across the GCC region. Journal of International Financial Markets, Institutions and Money, 25, 33-48. Balli, H. O., Balli, F., & Louis, R. J. (2013b). Time-varying spillover effects on sectoral equity Returns. International Review of Finance, 13(1), 67-91. Balli, F., Balli, H. O., & Luu, M. N. (2014). Diversification across ASEAN-wide sectoral and national equity returns. Economic Modelling, 41, 398-407. Balli, F., Basher, S. A., & Rana, F. (2014b). The determinants of the volatility of returns on cross-border asset holdings. Journal of International Money and Finance, 44, 1-23. Bekaert, G., Harvey, C. R. (1995). Time-varying world market integration. The Journal of Finance, 50(2), 403–444. Bekaert, G., Harvey, C. R. (1997). Emerging equity market volatility. Journal of Financial Economics, 43(1), 29–77.

13

ACCEPTED MANUSCRIPT Bekaert, G., Harvey, C. R. (2000). Foreign speculators and emerging equity markets. The Journal of Finance, 55(2), 565–613. Bekaert, G., Harvey, C. R., Lumsdaine, R. L. (2002). The dynamics of emerging market equity flows. Journal of International Money and Finance, 21(3), 295–350.

T

Bekaert, G., Harvey, C. R., Lundblad, C. (2005). Does financial liberalization spur growth? Journal of

RI P

Financial Economics, 77(1), 3–55.

Bodurtha, J. N., Cho, D. C., Senbet, L. W. (1989). Economic forces and the stock market: an international perspective. Global Finance Journal, 1(1), 21–46.

SC

Bracker, K., Docking, D. S., Koch, P. D. (1999). Economic determinants of evolution in international stock market integration. Journal of Empirical Finance, 6(1), 1–27.

Brooks, R., Del Negro, M. (2004). The rise in comovement across national stock markets: market

NU

integration or IT bubble? Journal of Empirical Finance, 11(5), 659–680. Campbell, J. Y., Hamao, Y. (1992). Predictable stock returns in the United States and Japan: a study of

MA

long-term capital market integration. The Journal of Finance, 47(1), 43–69. Caporale, G. M., Cipollini, A., Spagnolo, N. (2005). Testing for contagion: a conditional correlation analysis. Journal of Empirical Finance, 12(3), 476–489.

Caporale, G. M., Pittis, N., Spagnolo, N. (2006). Volatility transmission and financial crises. Journal of

ED

Economics and Finance, 30(3), 376–390. Carrieri, F., Errunza, V., Hogan, K. (2007). Characterizing world market integration through time. Journal

PT

of Financial and Quantitative Analysis, 42(4), 915. Cavaglia, S., Brightman, C., Aked, M. (2000). The increasing importance of industry factors. Financial Analysts Journal, 41–54.

CE

Chelley-Steeley, P. L. (2005). Modeling equity market integration using smooth transition analysis: a study of Eastern European stock markets. Journal of International Money and Finance, 24(5), 818–

AC

831.

Chen, G.-M., Firth, M., Meng Rui, O. (2002). Stock market linkages: evidence from Latin America. Journal of Banking & Finance, 26(6), 1113–1141. Chen, N.-F., Zhang, F. (1997). Correlations, trades and stock returns of the Pacific-Basin markets. PacificBasin Finance Journal, 5(5), 559–577. Diamandis, P. F. (2009). International stock market linkages: evidence from Latin America. Global Finance Journal, 20(1), 13–30. Diebold F.X., Yilmaz, K. (2009) Measuring financial asset return and volatility spillovers, with application to global equity markets. Economic Journal 119, 158–171. Engle, R., Gallo, G., Velucchi, M. (2008). A MEM-based analysis of volatility spillovers in East Asian financial markets. Available at SSRN 1283254. Floros, C. (2008). Modelling volatility using GARCH models: evidence from Egypt and Israel. Middle Eastern Finance and Economics, 2, 31–41. Goetzmann, W. N., Li, L., Rouwenhorst, K. G. (2005). Long-term global market correlations*. The Journal of Business, 78(1), 1–38.

14

ACCEPTED MANUSCRIPT Graham, M., Kiviaho, J., Nikkinen, J. (2012). Integration of 22 emerging stock markets: a three-dimensional analysis. Global Finance Journal, 23(1), 34–47. Hilliard, J. E. (1979). The relationship between equity indices on world exchanges. The Journal of Finance, 34(1), 103–114.

T

Jinjarak, Y., Wongswan, J., Zheng, H. (2011). International fund investment and local market returns.

RI P

Journal of Banking & Finance, 35(3), 572–587.

Khalifa, A., Hammoudeh, S., Otranto, E. (2012). Volatility Spillover, Interdependence, Comovements across GCC, Oil and US Markets and Portfolio Management Strategies in a Regime-Changing Environment.

SC

Centre for North South Economic Research, University of Cagliari and Sassari: Sardinia. Levine, R., Zervos, S. (1998). Capital control liberalization and stock market development. World Development, 26(7), 1169–1183.

NU

Ljung, G. M., Box, G. E. (1978). On a measure of lack of fit in time series models. Biometrika, 65(2), 297– 303.

MA

Lucey, B. M., Zhang, Q. (2010). Does cultural distance matter in international stock market comovement? Evidence from emerging economies around the world. Emerging Markets Review, 11(1), 62–78. Masih, A. M., Masih, R. (1999). Are Asian stock market fluctuations due mainly to intra-regional contagion effects? Evidence based on Asian emerging stock markets. Pacific-Basin Finance Journal, 7(3),

ED

251–282.

Neaime, S. (2002). Liberalization and financial integration of MENA stock markets. Paper presented at the

PT

ERF’s 9th annual conference on Finance and Banking. Neaime, S. (2006). Volatilities in emerging MENA stock markets. Thunderbird International Business Review, 48(4), 455–484.

CE

Neaime, S. (2012). The global financial crisis, financial linkages and correlations in returns and volatilities in emerging MENA stock markets. Emerging Markets Review, 13(3), 268–282.

AC

Ng, A. (2000). Volatility spillover effects from Japan and the US to the Pacific–Basin. Journal of International Money and Finance, 19(2), 207–233. Pretorius, E. (2002). Economic determinants of emerging stock market interdependence. Emerging Markets Review, 3(1), 84–105. Scheicher, M. (2001). The comovements of stock markets in Hungary, Poland and the Czech Republic. International Journal of Finance & Economics, 6(1), 27–39. Stiglitz, J. E. (2002). Information and the change in the paradigm in economics. The American Economic Review, 92(3), 460–501. Susmel, R. (2001). Extreme observations and diversification in Latin American emerging equity markets. Journal of International Money and Finance, 20(7), 971–986. Tay, N. S., Zhu, Z. (2000). Correlations in returns and volatilities in Pacific-Rim stock markets. Open Economies Review, 11(1), 27–47. Umutlu, M., Akdeniz, L., Altay-Salih, A. (2010). The degree of financial liberalization and aggregated stockreturn volatility in emerging markets. Journal of Banking & Finance, 34(3), 509–521.

15

ACCEPTED MANUSCRIPT Worthington, A., Higgs, H. (2004). Transmission of equity returns and volatility in Asian developed and emerging markets: a multivariate GARCH analysis. International Journal of Finance & Economics, 9(1), 71–80.

evidence. Journal of Applied Econometrics, 21(6), 727–744.

T

Yang, J., Hsiao, C., Li, Q., Wang, Z. (2006). The emerging market crisis and stock market linkages: further

RI P

Yilmaz, K. (2010). Return and volatility spillovers among the East Asian equity markets. Journal of Asian Economics, 21(3), 304–313.

Yu, J.-S., Hassan, M. K. (2008). Global and regional integration of the Middle East and North African

AC

CE

PT

ED

MA

NU

SC

(MENA) stock markets. The Quarterly Review of Economics and Finance, 48(3), 482–504.

16

ACCEPTED MANUSCRIPT T ABLES

Skew

Kurt

Q(1)

Asia Bangladesh India Indonesia South Korea Malaysia Pakistan Philippines Singapore Taiwan Thailand Vietnam

-0.19 0.13 0.02 0.06 0.08 0.02 0.06 0.09 -0.03 0.02 -0.12

3.6 4.41 6.21 5.2 4.12 4.61 4.4 3.29 4.31 5.23 5.27

-0.24 -0.29 -0.48 -0.51 -0.29 -1.00 -0.14 -0.34 -0.47 0.59 0.27

5.11 5.35 12.34 9.94 18.20 7.62 6.71 6.52 5.45 10.70 5.12

-0.07 -0.04 -0.02 -0.07** -0.01 0.10*** -0.05* 0.02 0.03 -0.02 0.22***

GCC Bahrain Jordan Kuwait Oman Qatar Saudi Arabia UAE

-0.46 0.03 0.00 0.24 0.06 -0.07 0.10

3.21 2.47 3.51 3.7 3.51 4.32 5.1

-1.39 -0.43 -0.85 -1.03 -0.03 -1.32 -0.65

East and Central Europe Bulgaria Czech Republic Estonia Hungary Latvia Lithuania Poland Romania Russia Turkey Ukraine

-0.32 0.14 0.22 0.16 0.04 -0.01 0.20 -0.05 0.21 0.06 -0.70

4.56 4.25 4.22 5.25 4.54 4.06 5.82 5.67 7.13 7.04 6.47

0.19 0.11 0.21 0.15 0.12

Latin America Argentina Brazil Chile Colombia Mexico Peru US Euro area Japan

SC

NU

Q'(1)

Q'(4)

-0.02 -0.03** 0.07*** -0.03*** -0.12*** 0.02*** -0.01*** 0.01 0.02** -0.01** 0.13***

0.20*** 0.22*** 0.26*** 0.31*** 0.19*** 0.30*** 0.22*** 0.29*** 0.23*** 0.10*** 0.34***

0.13*** 0.17*** 0.25*** 0.34*** 0.47*** 0.20*** 0.11*** 0.18*** 0.19*** 0.07*** 0.25***

0.05 0.00 0.08* 0.03 -0.01 0.03 0.04

0.09*** 0.05*** -0.05*** 0.05*** 0.07 -0.04 0.02**

0.21*** 0.25*** 0.32*** 0.13*** 0.30*** 0.19*** 0.08*

0.16*** 0.31*** 0.18*** 0.10*** 0.20*** 0.30*** 0.07***

-2.14 -0.05 -0.54 -1.05 -1.46 -0.06 -0.36 -0.93 -0.43 -0.37 0.13

16.77 12.34 7.50 13.03 38.79 13.90 6.84 7.98 8.80 4.99 12.03

0.12** -0.10*** -0.01 -0.02 0.02 0.07 -0.01 0.02 0.01 0.03 0.00

0.05*** -0.05*** 0.06* 0.00*** 0.05 0.07*** -0.05** 0.00*** -0.06*** 0.01* 0.01**

0.17*** 0.51*** 0.35*** 0.13*** -0.01 0.18*** 0.25*** 0.16*** 0.14*** 0.14*** 0.29***

0.11*** 0.17*** 0.17*** 0.05*** 0.02 0.09*** 0.28*** 0.09*** 0.29*** 0.12*** 0.09***

4.21 2.63 4.52 4.02 2.42

-0.55 -0.48 -0.41 -0.58 -0.15

5.55 7.69 7.41 7.72 8.29

0.05** -0.02 0.00 -0.08** -0.05

0.00*** 0.06* 0.11* -0.08*** 0.07

0.09*** 0.14*** 0.19*** 0.18*** 0.15***

0.11*** 0.08*** 0.33*** 0.16*** 0.14***

0.14 0.17 0.19 0.22 0.24 0.23

5.82 6.25 3.37 4.21 4.36 4.41

-0.11 -0.94 -0.32 -0.59 -0.55 -0.15

6.14 11.94 6.03 10.32 7.92 8.23

0.03 0.02 0.01 0.03 -0.02 -0.03

0.07** 0.00 0.01* 0.04*** -0.01** -0.03**

0.21*** 0.19*** 0.16*** 0.25*** 0.21*** 0.27***

0.30*** 0.15*** 0.21*** 0.12*** 0.21*** 0.37***

0.13 0.07 -0.03

2.33 3.47 3.19

-0.57 -0.32 -0.10

7.40 5.56 5.40

-0.10*** -0.12*** -0.06**

-0.04*** -0.03*** -0.02*

0.21*** 0.42*** 0.19***

0.19*** 0.21*** 0.04***

CE

PT

ED

13.35 9.19 6.85 12.69 7.33 8.22 9.53

AC

Africa Egypt Morocco Nigeria South Africa Tunisia

Q(4)

T

STD

RI P

Mean

MA

Table 1 – Stock Return Statistics

17

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA

NU

SC

RI P

T

The statistics are reported: mean, standard deviation (STD), skewness (Skew), kurtosis (Kurt), autocorrelations of order 1 and 4 (Q(1) and Q(4)), and autocorrelations of the squared time series of order 1 and 4 (Q†(1) and Q†(4)). *, **, and *** indicate that the Ljung and Box (1978) test statistic is significant at the 10%, 5%, and 1% levels, respectively.

18

ACCEPTED MANUSCRIPT Table 2 – Determinant Factor Statistics STD

Maximum (%)

Minimum (%)

Observations

Export to US Export to EU Export to Japan

15.3 24.17 0.95

16.11 20.4 1.2

79.96 73.76 5.58

1.98 1.38 0.05

39 39 39

Import to US Import to EU Import to Japan

11.16 24.77 0.71

10.83 17.03 0.68

55.11 62.28 2.45

1.47 5.37 0.03

39 39 39

Debt investments from US Debt investments from EU Debt investments from Japan

17.97 32.53 5.43

16.11 13.86 7.8

72.09 66.55 36.18

0.00 11.89 0.00

39 39 39

Equity investments from US Equity investments from EU Equity investments from Japan

29.19 16.17 0.82

20.96 11.34 1.02

68.95 52.37 4.18

0.00 0.00 0.00

39 39 39

Distance to US (km) Distance to EU (km) Distance to Japan (km)

9,279 5,672 9,174

16,180 12,098 18,550

3,369 546 1,157

39 39 39

RI P SC

NU

MA 3,219 3,725 4,045

T

Mean (%)

AC

CE

PT

ED

The table reports the summary statistics for the bilateral trades, security investments and distance between selected emerging markets and the economic powerhouses of the US, the Euro area (EU) area, and Japan.

19

ACCEPTED MANUSCRIPT Table 3 –Spillovers

PT

CE

AC

24.40 18.21 20.32 14.31 16.31 19.56 30.23

8.23 12.11 7.63 10.32 5.12 6.21 6.41

11.03 8.69 18.61 14.16 12.11 8.61 13.12

T

Japan

NU

9.61 18.94 17.71

7.129.92 9.51 6.51

12.61 13.01

18.31 20.02 21.14 18.96 28.67 18.95

14.05 19.42 14.24 16.19 13.52 15.65

2.45 14.31 4.50 12.61 3.61 10.61

12.41 13.56 13.53 11.20 12.45 13.80 11.67 11.78 12.34 14.16 12.59

16.34 16.62 13.71 15.61 15.36 18.01 19.61 15.16 18.43 17.33 10.56

4.13 2.89 4.20 4.37 5.21 9.20 3.41 3.71 9.92 3.24 3.61

21.35 16.45 13.11 30.97 18.23 15.11

11.21 8.84 22.13 15.33 16.33 17.41

3.61 2.01 2.05 5.48 5.35 2.12

16.63 15.82 19.55 16.52 21.34 20.12

12.15 10.75 10.21 10.34 7.45 10.65

2.87 5.61 2.33 7.85 6.61 4.56

MA

ED

Thailand Vietnam GCC Bahrain Kuwait Oman Qatar Saudi Arabia UAE East and Central Europe Bulgaria Czech Republic Estonia Hungary Latvia Lithuania Poland Romania Russia Turkey Ukraine Africa Egypt Jordan Morocco Nigeria South Africa Tunisia Latin America Argentina Brazil Chile Colombia Mexico Peru

Euro

RI P

Singapore

US

SC

Asia Bangladesh India Indonesia South Korea Malaysia Pakistan Philippines

The table reports the percentage variance ratios of spillover shocks for the constant spillover model from the US, the Euro area, and Japanese markets.

20

ACCEPTED MANUSCRIPT Table 3 a–Spillovers for period between 20082013 Japan

26.5

9.65

13.55

India Indonesia

26.54 25.32

13.65 6.33

11.21 20.21

South Korea Malaysia

19.32 18.51

11.21 6.55

15.61 13.54

Pakistan Philippines

24.51 36.21

1025 7.54

13.54 14.52

Singapore

12.12

10.24

Thailand

21.54

8.54

Vietnam GCC

22.21

10.54

Bahrain Kuwait

26.21 22.52

19.14 23.25

Oman Qatar

20.64 21.22

15.51 20.52

6.7 13.54

Saudi Arabia UAE

31.32 21.36

15.56 16.52

4.86 12.15

19.32

5.64

18.24 15.45

5.21 5.61

16.41 20.21

19.75 20.67

5.67 8.65

16.98 13.55

22.54 24.32

10.19 20.11

13.66 12.32

20.55 20.32

6.67 10.64

19.55 16.66

23.21 14.56

10.25 6.32

15.65

3.69

Hungary Latvia

Romania Russia Turkey Ukraine

AC

Lithuania Poland

CE

Africa Egypt

23.33

NU

11.26 12.01

MA

10.09

ED

PT

East and Central Europe 15.25 Bulgaria 14.14 Czech Repu. 12.32 Estonia

SC

Asia Bangladesh

T

Euro

RI P

US

4.31 16.63

Jordan Morocco

18.67 15.65

11.21 25.54

3.55 5.69

Nigeria South Africa

33.64 20.21

17.78 22.8

6.97 6.54

Tunisia Latin America

17.14

19.68

5.62

Argentina Brazil

18.36 16.32

18.65 14.01

5.21 6.66

Chile Colombia

21.38 18.84

12.12 16.35

3.54 10.41

21

ACCEPTED MANUSCRIPT Mexico

32.55

10.22

9.32

Peru

26.36

10.47

5.21

AC

CE

PT

ED

MA

NU

SC

RI P

T

The table reports the percentage variance ratios of spillover shocks for the constant spillover model from the US, the Euro area, and Japanese markets.

22

ACCEPTED MANUSCRIPT Table 4 – Cross-Sectional Analysis Euro area 41.36 (12.41)***

Japan 63.55 (23.33)***

Investment (INV)

34.32 (4.33)***

21.23 (9.11)**

2.14 (6.51)

Language (LANG)

6.21 (2.70)**

Capitalization (CAPT)

-1.61 (0.89)*

SC

RI P

T

USA 19.31 (6.03)***

Trade (TRADE)

-

NU

Colonial (COL)

-4.21 (4.55) 0.45 39

Distance (DIST)

MA

R2 Observations

2.12 (2.45)

-1.24 (0.68)* -0.44 (4.02) -2.34 (1.23)* 0.43 39

-

0.23 (0.71) -

-13.45 (4.52)** 0.36 39

AC

CE

PT

ED

The table reports the cross-sectional estimation results with standard errors in parentheses for each of the emerging markets following the formula:  ୨ = α଴ + αଵ  ௝ + αଶ  ୨ + αଷ  ୨ + αସ ୧ + αହ  ୨ + α଺   ୨ + ε୧ ,  ୧ ௜ ୧ ୧ ୧ ୧ See the text for the definition of the variables. *, **, and *** indicate that the t-statistics are significant at the 10%, 5%, and 1% levels respectively.

23

ACCEPTED MANUSCRIPT Table 5 – Aggregate Cross-Sectional Analysis (1) 28.72 (6.75)***

Trade (TRADE)

(2)

(3)

(4)

SC NU

Colonial (COL)

0.38 117

0.29 117

0.09 117

MA

Distance (DIST) R2 Observations

8.66 (3.55)**

-2.45 (0.75)**

Capitalization (CAPT)

0.07 117

-2.13 (0.85)**

2.45 (1.78)

0.05 117

(6) 26.13 (8.51)*** 35.55 (2.34)***

RI P

10.34 (3.45)***

Language (LANG)

(5)

T

33.41 (6.87)***

Investment (INV)

(4)

2.72 (1.45)*

-1.87 (0.65)** 0.05 117

-1.82 (1.05)* 0.61 117

AC

CE

PT

ED

The table reports the cross-sectional estimation results with standard errors in parentheses for each of the emerging markets following the formula:  ୨ = α଴ + αଵ  ௝ + αଶ  ୨ + αଷ  ୨ + αସ ୧ + αହ  ୨ + α଺   ୨ + ε୧ ,  ୧ ௜ ୧ ୧ ୧ ୧ See the text for the definition of the variables. *, **, and *** indicate that the t-statistics are significant at the 10%, 5%, and 1% levels respectively.

24

ACCEPTED MANUSCRIPT F IGURES 2. Export Percentage

60%

40% 35% 30% 25% 20% 15% 10% 5% 0%

30%

US

20% Euro

10% Japan

US Euro Japan

NU

0%

RI P

40%

SC

50%

T

1. Security Investment Percentage

MA

Figure 1a: Security investment from developed markets to emerging Asian markets

30%

Figure 2a: Export to developed markets from emerging Asian markets

20%

ED

25% 20% 15%

15% 10%

US

10%

PT

Euro

5%

Japan

0%

Figure 2b: Export to developed markets from emerging GCC markets

AC

Figure 1b: Security investment from developed markets to emerging GCC markets

50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0%

Euro

5%

Japan

CE

0%

US

80% 70% 60% 50% 40% 30% 20% 10% 0%

US Euro Japan

Figure 1c: Security investment from developed markets to emerging Eastern and Central European markets

1. Security Investment Percentage (cont.)

Euro Japan

Figure 2c: Export to developed markets from emerging Eastern and Central European markets

2. Export Percentage (cont.)

25

US

ACCEPTED MANUSCRIPT

80% 70% 60% 50% 40% 30% 20% 10% 0%

30%

US

20%

Euro

10%

Japan

0%

US

T

40%

RI P

50%

Euro Japan

SC

60%

Figure 2d: Export to developed markets from emerging African markets

NU

Figure 1d: Security investment from developed markets to emerging African markets

90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

MA

80% 70% 60% 50% 40% 30% 20% 10% 0%

US

Euro

PT

ED

Japan

CE

Figure 1e: Security investment from developed markets to emerging Latin American markets

US Euro Japan

Figure 2e: Export to developed markets from emerging Latin American markets

AC

The graphs visually compare the figures (in percent) for the determinants of security investments (1) and exports (2) between the emerging and the developed markets of the US, the Euro area, and Japanese markets, grouped according to the different regions of the emerging markets: Asia, Gulf Cooperation Council (GCC), Eastern and Central Europe, Africa, and Latin America. UAE, United Arab Emirates.

26