Culture and stock price synchronicity in emerging countries Daming Su University of Groningen Faculty of Economy and Business Master of Science, Business and Administration of Finance

(1)

Culture and stock price synchronicity in emerging countries

Daming Su

University of Groningen Faculty of Economy and Business

Master of Science, Business and Administration of Finance

Student number: S1939556

Supervisor: Dr. Auke Plantinga

Abstract

Stock prices tend to move collectively in emerging countries. In this paper, I explore the question what the determinants are for the high levels of stock price synchronicity in low income countries. I use three groups of variables, macro-economic fundamentals, protection of private property rights, and cultural dimension values, to explain this phenomenon. This paper complements prior research with the impact of cultural dimensions on stock price synchronicity and new proxies for macro-economic fundamentals. I find that Power Distance and geographic surface can explain why stock price movements exhibit a more synchronous manner in low income countries, after reducing the multicollinearity problems. Furthermore, their effects on stock price synchronicity persist over time.

JEL classification: G12; G14; G15; G38; O11; N20

(2)

1: Introduction

Stock price synchronicity measures the fraction of co-movements of stock prices in one market, in contrast to idiosyncratic movements of stock prices. It is often used as the measurement for the amount of firm-specific information that is capitalized into stock price1. Consistent with prior research, the simple correlation results show that stock prices tend to move up and down collectively in the emerging countries. In this paper, I address the question whether culture can explain the higher stock price synchronicity in the emerging countries.

Standard asset pricing models suggest that two types of risk determine stock returns, market risk and firm-specific risk, which in turn determines the co-movements and firm specific co-movements of stock prices in the market. As Roll (1988) suggests, market wide news can only explain a small fraction of stock returns, which implies stock returns also depend on the amount firm-specific information that is capitalized into stock prices. Piotroski and Roulstone (2004) also find that stock prices movements reflect the amount of firm-specific information. Consistent with these findings, a body of research argues that the information environment can influence stock price synchronicity or firm-specific stock price variations. Information environment refers to the ratio and rate of market wide news and firm-specific information that is incorporated into the stock prices. If market wide information dominates the stock market, stock price movements will show a synchronous manner. Prior research studies the determinants of stock price synchronicity by looking at factors that influence accounting information. For example, Morck, Yeung, and Yu (2000) explain that high stock price synchronicity is due to poor protection of the private property rights, which discourage private information from being incorporated into stock prices. Jin and Myers (2006) find that stock price synchronicity is a negative function of the degree of accounting transparency in the market. Gul, Kim, and Qiu (2010) studies the impacts of ownership concentration and audit quality on stock price synchronicity in China and argues that ownership concentration and auditor quality are negatively associated with stock price synchronicity. These arguments are similar in that these factors influence the degree of stock price

1

(4)

movements through their impacts on the accounting information disclosures. In this context, the influence of culture is relatively neglected. In this paper, I contribute to the determinants of stock price synchronicity by studying the impacts of culture.

Morck, Yeung, and Yu (2000) are the first to study stock price synchronicity across countries. Using data from 40 countries in 1995, they find that emerging countries generally exhibit a high degree of stock price synchronicity. They argue that it is attributed to the poor protections of the private property rights by governments rather than correlated economic fundamentals. Poor protections of the private property rights restrain the fraction of risk arbitrage trading in the market and consequently discourage information from being incorporated into share prices. A lack of risk arbitrages is accompanied by market wide stock price swings due to noise trading. However, the significance of the private property rights protections in their model does not mean that their explanations are complete. Their selection of measurement for economic fundamentals is open to doubt, since they present few literatures to support the association of economic fundamentals with stock price synchronicity. Other authors2, using various macro-economic fundamentals, have provided evidence that these effects of real economy are transmitted into financial markets (e.g. through the impacts on firms’ cash flows and discount rate) and cause market wide stock price variations. In this paper, I also contribute to this topic by adding new proxy for economic fundamentals, beside the cultural impacts.

I organize this paper as follows. In the next section, I describe the measurement of stock price synchronicity. In section 3, I present the literature reviews. In section 4, I present the summary of statistics and the simple correlation results. In section 5, I describe the methodology implemented in this paper. In section 6, I give the results and discussions of regressions. In section 7, I make the conclusion.

(5)

2: Measurement of stock price synchronicity

I measure the stock price synchronicity using the same construction in Morck, Yeung, and Yu (2000), the weekly average of the portion of stocks whose prices move in the same direction. It measures the major trend of stock price movements. For example, suppose that the stock market has 10 stocks, 5 stocks prices increase and 4 stocks prices decline and 1 stock price stays the same during a week. Then the stock price synchronicity is 5/9, where 5 indicates the number of stock prices that follow the major trend of stock price movements and 9 indicates the number of the changing stock prices in total. The construction of stock price synchronicity is as follows,

[

]

[

]

∑

₊ = t down jt up jt down jt up jt T j n n sum n n T S , 1 max , (1)

In equation (1), nupjt denotes the number of stocks whose value increases, while

down jt n denotes the number of stocks whose value declines in country j over the week t. I exclude zero returns stocks to avoid a potential bias due to non-trading stocks. T is the number of weeks applied. The value of S_j_,_Tis limited by the calculation of stock price synchronicity and its value is limited from 0.5 to 1. In order to enlarge the range of the stock price synchronicity value, I convert it into the adjusted measurement of synchronicity using natural logarithm function, presented in equation (2).

t j t j t j S S S , , * , 1 5 . 0 log − − = ₍₂₎

Where S_j_,_T is the measurement of synchronicity in equation (1). I lay out the summary of data statistics in Table 1 in section 4.

(6)

influence stock price movements. In the next section, I begin to present the factors that may influence stock price synchronicity.

3: Literature review

From the multivariate and correlation results, I observe a negative relationship between stock price synchronicity and per capita GDP at the significance level of 0.01. In other words, one can say stock price movements exhibit a more synchronous manner in the emerging countries or the low income countries. I develop three hypotheses to explain this phenomenon. Firstly, I hypothesize that stock price co-movements result from the synchronous co-movements of economic fundamentals. Chen, Roll and Ross, (1986), Wongbangpo and Sharma, (2002), and Abugri, (2008) have suggested that macroeconomic fundamentals can influence stock prices. They argue that macroeconomic fundamentals impact the firms’ future cash flow and the discount rate in a standard stock price valuation model. The development of the economy can improve the investors’ participation in the stock markets. I expect that the changes of macroeconomic fundamentals will be transmitted into the stock market. Secondly, synchronicity reflects the degree of governments’ respect for the private property rights. Morck, Yeung, and Yu (2000) argue that high stock price synchronicity is due to the poor protection of private property rights which in turn determine the dominance of market level information over firm level information. Thirdly, culture dimensions influence the stock price synchronicity through its impacts on the development of accounting practices and disclosure standards.

3.1 Macro-economic fundamentals and stock price synchronicity

(7)

3.1.1: Economy size

The size of economy may affect all the stock price movements in the market. Chen, Roll and Ross (1986), Fama (1981) and Mukhejee and Naka (1995) provide evidence that the stock prices are positively related to GNP. Wongbangpo and Sharma (2002) argue that the level of national output causes an overall effect on firms’ profitability, which is a factor in the evaluation model of stock prices. Since GNP influences the whole financial market, it causes an overall impact on the movements of stock prices. Furthermore, the size of economy also indicates the sensitivity of the real economy to the international events. Graham and Dennis (2010) argue that small economies tend to be heavily dependent on the international economy since both the range and the scale of products are limited and the customers find it relatively difficult to switch from domestic preferences to foreign goods. Small economy is more exposed to its export or import partners. Therefore, small economy often exhibit market wide variations of stock prices

I measure the size of economy by the natural logarithm of Gross Domestic Products (GDP) in million US dollars in the country. For the robustness test, I use the natural logarithm of Gross National Income (GNI) in million US dollars as an alternative measurement.

3.1.2: Country size

(8)

I measure the country size attribute by the natural logarithm of geographic size, in kilometer square. Regarding the robustness measurement, I use the natural logarithm of population size, in millions, as an alternative measure.

3.1.3: Economic growth

Economic growth is market wide information and can cause overall increase of stock prices. For example, Chen (1986) suggests that economic growth is a proxy for the business condition and is negatively associated with the expected stock returns. Wongbangpo and Sharma (2002) documents that the growth of stock market capitalization is often associated with the economic growth in ASEAN countries. This may in turn have a negative impact on most of the stock prices. This is consistent with the predictions in Chen, Roll, and Ross (1986) that stock market may attract more investors during the economy expansion, since one country’s economic growth can increase the investors’ money supply and encourage people to participate in the stock market.

As an inverse proxy for the economic growth, inflated economy increases the costs of firms’ input, decreases corporate profitability and worsens the expected future cash flow (Wongbangpo and Sharma, 2002). The economic inflation can cause an overall decrease of the value of stock prices, which are the present value of future cash flows. To capture the relationship between economic growth and stock price synchronicity, I use GDP growth rate as the proxy for the economic growth attribute. As an alternative, I use inflation rate as the inverse proxy for economic growth attribute.

3.1.4: Market competition

(9)

with many other corporations, the successions become industry wide or market wide news and have an impact on other firms’ stock price. In contrast, if there are many firms in the market which compete severely with each other and suppliers can easily find alternative customers, then news about a particular firm may cause little influence on the whole stock market.

To capture of the effect of market competition, I construct a Herfindahl index (Hjin equation (3)) to measure the degree of competition in one market as follows,

∑

∈ = 2 , ,t i jt j h H ₍₃₎

where hjis the weight of sales of each firm in country j over time period of t . High

value of Herfindahl index indicates lower intensity of competition in the market. Since this is the only indicator for market competition, I assume that it captures the effect of market completion completely.

3.1.5: Liquidity

(10)

measurement is that it cannot distinguish whether the cause of observations of zero returns is illiquidity or non-trading days. Zero return is calculated

as shown in the equation (4)

t j zero t j T j N n T ZR , , , * 1 = ₍₄₎

In equation (4), T denotes the number of periods applied. N is the total number of _j stocks in country j over time period of t . n_jzero is the observed number of zero returns in country j over time period of t.j . Zero return is an inverse proxy for liquidity attribute. I use the natural logarithm of weekly average of the turnover volume as an alternative measure for the liquidity attribute.

3.2: Protection of the private property rights and the stock price synchronicity

Synchronicity may be related to the quality of the shareholder protection rights. A number of authors suggest that poor investor protection is associated with higher correlation of stock returns. Aggarwal and Goodell (2008) find that emerging countries are often associated with higher stock price synchronicity and worse shareholder protection. Morck, Yeung, and Yu (2000) are the first to provide direct evidence about stock price synchronicity and protection of the private property rights. They argue that high stock price synchronicity is due to poor protection of the private property rights which discourages the amount of informed trading, leaving the dominance of noise over the market. If noise traders control the stock market, stock prices tend to move in a synchronous manner. Furthermore, stock markets with poor protection of the private property rights are exposed to rumors of political events, since their governments are often relatively unstable. One example provided by Fisman (2001) shows that more than 25% of Indonesian stocks are affected by the rumors of President Suharto's health. Because of the reasons above, one may expect that poor protection of the private property is negatively associated with stock price synchronicity.

(11)

corruption index, expropriation of private property, and repudiation index to describe governments’ respect for private property.

As described by La Porta et al. (1998), the corruption index measures corruption in government and is provided by International Country Risk (ICR). Low scores of this index indicates that one has to pay special payments to high government officials and pay additional payments to low government officials who in turn are connected with import and export licenses, exchange controls, tax assessments, policy protections, or loans. The expropriation index captures the risk of “outright confiscation or forced nationalization” in the country. The repudiation of contracts by government index is ICR's assessment of the risk of a modification in a contract taking the form of repudiation, postponement, or scaling down due to budget cutbacks, indigenization pressure, a change in government, or a change in government economic and social priorities.’ These three indexes is collected and built from 1982 to 1995, scaled from 0 to 10. I calculate government index by the natural logarithm of the sum of three indexes in each market. Low score of good government index indicates poor protection of the private property rights.

3.3: Culture and stock price synchronicity

Roll (1988) suggests that expected stock returns depend on two sources of risk: systematic risk and specific risk. He finds that capitalization of proprietary firm-specific information explains most of stock returns. Following his finding, Morck, Yeung, and Yu (2000, p. 258) argues that high stock synchronicity is due to limited incorporation of firm-specific information into stock prices and find that it is poor protection of the private property rights that discourage the capitalization of the firm-specific information. Other authors3 also study stock price synchronicity using various factors that influence accounting information. In this context, the impact of culture on synchronicity does not perceive deserved appreciation, although previous literatures, like Gray (1988), have built the model and argue that the development of accounting

(12)

practices and standards reflects the cultural dimensions. In this section, I review previous researches on the relationship between culture and accounting secrecy. Using this relationship as a link, I introduce my predictions of cultural impact on stock price synchronicity.

In 1988, Gray (1988, p. 7) proposes a theoretical model about how culture and societal values can influence accounting sub-culture (p. 7). In his model, he uses four dimensions of culture values4 from Hofstede (1984) as the proxies for culture and creates accounting values inspired by previous literatures as the proxy for accounting sub-culture values. One of his predictions is that culture values can determine accounting secrecy or transparency. Salter and Niswander (1995, p. 394) find that Gray’s model has reasonable predicted power for the development of accounting practices. Zarzeski (1996) also provides empirical evidence supporting cultural values have a significant impact on financial disclosures. Therefore, culture is closely related to the development of information disclosure practices across countries.

Gray (1988, p. 11) hypothesizes that “ the higher a country ranks in terms of uncertainty avoidance and power distance and the lower it ranks in terms of individualism and masculinity, then the more likely it is to rank highly in terms of secrecy.” He defines secrecy as “a preference for confidentiality and the restriction of disclosure of information about the business only to those who are closely involved with its management and financing as opposed to a more transparent, open and publicly accountable approach (Gray, 1988, p. 8). A preference for uncertainty avoidance, which is the degree to which people feel uncomfortable about uncertainty or ambiguity, encourages secretive or confidential accounting practices, since it reduces potential conflicts. A secretive society is consistent with high power distance since it discourages information sharing and leads to information inequality. Collectivism, as opposed to individualism, means a tight knit social framework where people exchange information only with closely related groups. It is negatively associated with accounting transparency. Although being less important, a preference for masculinity encourages its people to achieve material success and heroism and therefore people are more open to disclosed information.

Since more transparent approach indicates more information available to public investors, I expect that higher transparency is associated with lower stock price

(13)

synchronicity. If cultural dimensions have significant influence on accounting secrecy or transparency, they should also have influence on stock price synchronicity. Therefore, I expect stock price synchronicity shows up in countries with lower values of uncertainty avoidance and power distance and higher values of individualism and masculinity.

To capture the relationship between culture and stock price synchronicity, I use cultural values from Hofstede (1984). According to Hofstede (1984, p83-84), the concept of Individualism (IDV) is the extent to which individuals depend on each other. He defines that individuals in an Individualism culture only care about themselves and are supposed to express their characteristics freely. On the contrary, people in Collectivism culture prefer a tightly knit social framework where individuals all serve the goals of the groups. Power Distance (PD) is the degree to which people deal with inequalities in organizations. A larger Power Distance encourages a hierarchical structure where the positions set the power to people and people have larger power inequalities. People in the higher hierarchy are likely to maintain mysterious to lower hierarchy and constrain secrets within their positions. Uncertainty Avoidance (UA) describes the degree people can put up with ambiguity and uncertainty. In a strong Uncertainty Avoidance society, people are likely to stick to rigid rules or codes. On the contrary, people in a weak Uncertainty Avoidance society, people are likely to relax the rules and handle with problems based on the practical situations. Masculinity (MAS) indicates that the society promote heroism, assertiveness, and material success.

4: Data

(14)

World Bank’s World Development Indicators. For the country-level economic fundamentals, such as GDP and GDP growth rate, I collect from IMF, World Bank, and CIA factbook. In addition, I collect the index of good government from La Porta (1988).

Among those data collected above, I find some of them highly skewed and I use natural logarithm to adjust for this problem. In particular, I find that the difference between maximum and mean value lead to highly skewed distribution of GDP, GNI, GEO, POP, GRO, INF, TUR, and per capita GDP. After adjustment, the summary of statistics is as follows,

Table 1 Summary statistics

Variables Mean Median Maximum Minimum Std. Dev.

SYN 0.66 0.66 0.80 0.57 0.04 AD_SYN -0.76 -0.79 0.70 -1.86 0.44 Economic fundamentals GDP 6.26 5.94 9.57 4.37 1.16 GNI 13.15 12.85 16.48 11.21 1.20 GEO 12.63 12.76 16.12 6.15 2.32 POP 17.32 17.24 21.01 15.23 1.44 GRO 2.87 2.93 14.20 -8.02 3.80 INF 3.38 2.68 20.78 -1.71 2.89 HERF 0.09 0.04 0.99 0.00 0.17 ZR 0.19 0.14 0.62 0.01 0.15 TUR 12.34 12.24 16.11 9.33 1.42 GDP/p 9.86 10.26 10.88 7.67 0.85 GOV 23.09 25.13 29.59 6.53 6.09 Culture UAI 62.43 64.50 112.00 8.00 24.50 IDV 48.83 49.50 91.00 13.00 25.29 PD 55.25 59.00 104.00 11.00 20.53 MAS 50.23 53.00 95.00 5.00 18.95

SYN is calculated as in equation (1). AD_SYN is another measurement of synchronicity in equation (2). GDP is the natural logarithm of gross domestic product (GDP) in million US dollar. GNI is natural logarithm of gross national income (GNI) in million US dollars. GEO is the natural logarithm of geography size in one country. POP is the natural logarithm of population. GRO is annual GDP growth rate. INF is annual inflation rate. HERF is short for the Herfindahl index, which is calculated from equation (3).ZR is the proportion of zero returns in one market, which is calculated using equation (4). TUR is natural logarithm of the weekly average of turnover volume. GDP/p is short for per capita GDP in each country. GOV is short for index of good government, calculated by the sum of three indexes from La Porta (1988), corruption index, expropriation, and repudiation of contracts by governments. The four cultural dimension values are Uncertainty Avoidance Index (UAI), Individualism (IDV), Power Distance (PD), and Masculinity (MAS).

(15)

Looking at correlation results in table 2, I observe correlations between dependent variables and independent variables, and also correlations between independent variables themselves in equation (5). Stock price synchronicity and adjusted synchronicity is positively related at significant level of 0.01. I also observe that some of independent variables are correlated with each other. For example, GDP is significant correlated with geographic size, the Herfindahl index and zero returns. If I include them within one multivariate regression (5) at the same time, this may lead to bias caused by multicollinearity. I solve this problem by running regression tests on each individual variable. If the results of individual regressions differ from those from the multivariate regressions, I prefer the results from individual regressions to reduce bias caused by multicollinearity. Furthermore, I also run regressions on alternative measurements of macro-economic fundamentals to reduce potential multicollinearity caused by the choice of measurements.

(16)

may raise multicollinearity problem in the multivariate regression in equation (5) below.

(17)

(18)

5: Methodology

Consistent with the findings of Morck, Yeung, and Yu (2000), I observe that stock prices exhibit more synchronous movements in the low income countries. In order to explain this phenomenon, I start to search for the channels that may explain the relatively high stock price synchronicity in the low income countries. I run five panel regressions including three groups of variables, which are macro-economic fundamentals, the protection of the private property rights and the cultural dimension values.

I include economic variables in all the 3 regressions and in the latter regressions as control variables. This model allows me to estimate the ability of economic fundamentals, private property protections, and culture to explain the high synchronicity in the low income countries. If the coefficient of culture values is significant in Equation (5) below, I conclude that culture dimensions can explain stock price synchronicity across countries. If it can render GDP/p insignificant in Equation (5), it indicates that it explain the relative high stock price synchronicity in the low income countries.

ε

β

α

+ + + + + = jT jT jT jT T j GDP p X GOV RT S*_, ₁* / _, ₂* _, ₃* _, ₄* _, (5) where * ,T j

S represents the adjusted measurement of synchronicity in country j in equation (2) . GDP/pj,T is the per capita GDP in country j . X_j_,_T is the vector of economic fundamentals. GOVj,Tis the level of protection of private property rights in country j. RTj,T represents cultural dimensions in country j .

In order to overcome the problem of multicollinearity, I run all the dependent variables individually as in the Equation (6) below and compare them with the results of Equation (5). If the coefficient of any variable is insignificant, I drop it out the Equation (6) to reduce the multicolinearity problems.

ε

β

α

+

=

_j _j j

GDP

p

IND

S

* ₁

*

/

₂

*

₍₆₎

Where GDP/pjis per capita GDP in country j .I run regression test on one of the

(19)

6: Results and Discussions

6.1: Regressions on individual variables

From the multivariate regressions using all the macro-economic fundamentals and cultural dimension values (not shown), the values of Durbin-Watson test are around 0.5, which indicates there are multicollinearity problems. In order to reduce this problem, I run each of dependent variables individually as in the Equation (6). Looking at table 3, I find that UAI, IDV, and GRO are not related to AD_SYN at all and thereby I drop them out of the multivariate regressions. Three dependent variables, say GOV, PD, and GEO, render GDP/P insignificant, which means these factors explain the high stock synchronicity in the low income countries. In particular, PD and GEO is significantly positively related to AD-SYN, GOV is negatively related to AD_SYN. Using alternative measurements for macro-economic fundamentals, I find similar results to those in table 3.

Table 3

Individual impacts on stock price synchronicity

Dependent Variable:

AD_SYN GOV UAI IDV PD MAS GDP GEO GRO HERF ZR

Does it significantly related to

AD_SYN ? Yes No No Yes Yes Yes Yes No Yes Yes Does it render the

coefficient of GDP/p insignificant in

equation (6)? Yes No No Yes No No Yes No No No AD_SYN is another measurement of synchronicity in equation (2). GOV is short for index of good government, calculated by the sum of three indexes from La Porta (1988), corruption index, expropriation, and repudiation of contracts by governments. The four cultural dimension values are Uncertainty Avoidance Index (UAI), Individualism (IDV), Power Distance (PD), and Masculinity (MAS). GDP is the natural logarithm of Gross Domestic Product (GDP) in million US dollar. GEO is the natural logarithm of geography size in one country. GRO is annual growth rate of GDP. HERF is short for the Herfindahl index, which is calculated from equation (3). ZR is the proportion of zero returns in one market, which is calculated using equation (4). GDP/p is short for per capita GDP in each country. Refer detailed signs of significant variables to the figure 5 in the appendix.

(20)

6.2: Macro-economic fundamentals

Looking at panel A in Table 4, I observe that 3 macro-economic variables, GDP,

Herfindahl index, and zero returns are significantly related to stock price synchronicity at 0.05 level. In particular, AD_SYN are positively related to GDP and is negatively related to HERF and ZR. The positive impact of GDP indicates that stock prices tend to move together in a large economy and it is against my expectations. Since Herfindahl index is an inverse proxy for the intensity of market competition, it implies that stock price synchronicity increases when the intensity of market competition decrease. It confirms my expectation that if the market is dominated by a few large companies, firm-specific news may cause market level stock price movements. ZR is an inverse proxy for stock liquidity and the result is consistent with the idea that highly liquid markets facilitate the capitalization of firm specific information into stock prices and thereby exhibit more idiosyncratic stock price movements.

(21)

6.3: Protection of private property rights

I add one more variable into the multivariate regression in Equation (5) to capture the effect of protection of the private property rights. Looking at panel B in Table 2, I observe that government index has a significant negative effect on the stock price synchronicity. This implies that if one government offers good quality of private property protection, then stock prices move in a diversified way. This is consistent with the idea that good protection of the private property facilitates informed trading transactions, decreases market noise trading and thereby associates with low degree of stock price synchronicity.

Furthermore, GOV render GDP/p insignificant, which indicates that the quality of government protection of the private property rights can explain the relatively high stock price synchronicity in the emerging countries. Furthermore, , the significant effects of economic fundamentals in Panel B differ little from the results in Panel A, which indicates that their effects are independent of the effect of GOV.

6.4: Culture

(22)

Table 4

Panel regression results

Dependent Variable:

AD_SYN PD MAS GOV GDP/p GDP GEO HERF ZR

Panel A Coefficient -0.150 0.074 0.010 0.563 -0.500 (Probability) (0.001) (0.038) (0.575) (0.002) (0.040) Panel B Coefficient -0.0321 -0.013 0.079 0.010 0.393 -0.827 (Probability) (0.000) (0.807) (0.017) (0.545) (0.018) (0.001) Panel C Coefficient 0.000 -0.031 -0.014 0.073 0.012 0.397 -0.834 (Probability) (0.797) (0.000) (0.791) (0.061) (0.513) (0.018) (0.001) Panel D Coefficient 0.000 -0.032 -0.014 0.079 0.010 0.392 -0.828 (Probability) (0.942) (0.000) (0.801) (0.018) (0.546) (0.019) (0.001) Panel E Coefficient 0.000 0.000 -0.031 -0.015 0.073 0.012 0.397 -0.835 (Probability) (0.931) (0.794) (0.000) (0.784) (0.062) (0.513) (0.018) (0.001)

This tables show the results of five panel regressions. I mark the significant coefficients with bold letters. AD_SYN is another measurement of synchronicity in equation (2). PD is short for Power Distance, and MAS is short for Masculinity. GOV is short for index of good government, calculated by the sum of three indexes from La Porta (1988), corruption index, expropriation, and repudiation of contracts by governments. GDP/p is short for per capita GDP in each country. GDP is the natural logarithm of gross domestic product (GDP) in million US dollar. GEO is the natural logarithm of geography size in one country. HERF is short for the

(23)

6.5: Robustness test

In order to test the robustness of the above, I use alternative measurements for macro-economic variables. In particular, I replace GDP with GNI to describe economy size, replace GEO with POP to describe geography size, replace GRO with INF to describe economy growth and replace ZR with TUR to measure market liquidity. In order to avoid the problem of multicollinearity, I run each of these variables individually like the results shown in Table 3. The results are presented in Table 5. Comparing the results in Table 3 with those in Table 5, the effects of alternative measurements are similar. Therefore, the panel regressions using macro-economic variables are robust.

Table 5

Individual regressions of alternative measurements

Dependent Variable:

AD_SYN GNI POP INF TUR

Does it significantly related to AD_SYN ? Yes Yes No Yes Does it render the coefficient of GDP/p

insignificant in equation (6)? No Yes No No

This table shows the regression tests on each individual variable using the alternative measurement for macro-economic fundamentals. AD_SYN is another measurement of synchronicity in equation (2). GDP/p is short for per capita GDP in each country. GNI is natural logarithm of gross national income (GNI) in million US dollars. POP is the natural logarithm of population. INF is annual inflation rate. TUR is natural logarithm of the weekly average of turnover volume.

6.6: Time period effects

(24)

Table 6 Time period effects

Dependent Variable:

AD_SYN PD MAS GOV GDP/p GDP GEO HERF ZR

Coefficient 0.000 0.001 -0.028 -0.038 0.044 0.018 0.389 -0.953

(Probability) 0.856 0.443 (0.000) 0.455 0.237 0.291 (0.015) (0.000)

AD_SYN is another measurement of synchronicity in equation (2). PD is short for Power Distance, and MAS is short for Masculinity. GOV is short for index of good government, calculated by the sum of three indexes from La Porta (1988), corruption index, expropriation, and repudiation of contracts by governments. GDP/p is short for per capita GDP in each country. GDP is the natural logarithm of gross domestic product (GDP) in million US dollar. GEO is the natural logarithm of geography size in one country. HERF is short for the Herfindahl index, which is calculated from equation (3). ZR is the proportion of zero returns in one market, which is calculated using equation (4).

7: Conclusion

Consistent with the previous findings, I find evidence that the stock price synchronicity is higher in the low income countries. In other words, stock prices tend to move together in the emerging countries. While previous researches have shown that there are different factors that can influence the stock price synchronicity through their impacts on the firm-specific information disclosure, the impacts of culture are relatively neglected in this context. Based on the predictions of Gray (1988)’s model, I introduce cultural dimension values to explain the stock price synchronicity. I use three groups of variables in the multivariate regression equation, which are macro-economic variables, protection of the private property rights, and the culture dimension values. I have found that GOV and 3 economic fundamentals, GEO, HERF, and ZR, are significantly related with the stock price synchronicity. In order to reduce multicollinearity, I run regressions on each individual variable. From the individual results of each variable in Table 3 and multivariate regressions in Table 4, I find that 3 variables, GOV, PD, and GEO render GDP/p insignificant in the multivariate regression. This means each of the 3 variables can explain the high stock price synchronicity in the emerging countries. Other factors that influence AD_SYN include MAS, GDP, HERF, and ZR.

(25)

These findings indicate that stock prices in the low income countries reflect more market wide information. The determinants of that stock price co-movements shows that the stock price synchronicity tend to be higher in a country with a poor protection of the private property rights, a large power distance, a large geographic surface, and a preference for masculinity, economic growth, less competition, less liquidity. The effect of PD on the stock price synchronicity is consistent with its influence on accounting transparency. A preference for PD discourages information sharing and leads to information inequality. Therefore, the stock price synchronicity increases with a larger Power Distance. The effects of GOV and economic fundamentals are robust to alternative measurements of the same attribute. Their effects also persist over the time period from 2005 to 2009. These findings implicitly benefit the international investors as they select asynchronous stock prices across countries or estimate the predictability of stock returns.

However, I should mention that I have to apply the findings about cultural dimension values and the stock price synchronicity with caution. The validity of the findings is limited in that the cultural dimension values are outdated, as Hofstede (1980) build the index from 1967 to 1973. The impact of culture on the stock price synchronicity requires an updated proxy for culture. Furthermore, this paper studies the determinants of stock price synchronicity by looking at the factors that influence the capitalization of market wide information and firm specific information. I use macro-economic fundamentals to capture the impact of market wide information and implicitly assume the rest of the impact comes from firm-specific information. Further studies should create a proxy for firm specific information, so as to its study the complete impact on the stock price synchronicity.

(26)

References

[1] Abugri, B.A., 2008. Empirical relationship between macroeconomic volatility and stock returns: Evidence from Latin American markets. International Review of

Financial Analysis 17, 396-410.

[2] Aggarwal, R., and Goodell, J.W., 2008. Equity premia in emerging markets: National characteristics as determinants. Journal of Multinational Financial

Management 18, 389-404.

[3] Chen, N., and Hsieh, D., 1985. An exploratory investigation of the firm size effect.

Journal of Financial Economics 14, 451-471.

[4] Chen, N.F., 1991. Financial investment opportunity and the

macroeconomy. Journal of Business 46, 529–554.

[5] Chen, N.F., Roll, R., and Ross, S.A., 1986. Economic forces and stock market.

Journal of Business 59(3), 383-403.

[6] De Long, J.B., Shleifer, A., Summers, L.H., Waldmann, R.J., 1990. Noise trader risk in financial markets. Journal of Political Economy 98, 703-738.

[7] Fama, E.F, 1981. Stock returns, real activity, inflation, and money. American

Economic Review 71, 545–565.

[8] Fan, J. P.H., Rui, O.M., and Zhao, M.X., 2008. Public governance and corporate finance: Evidence from corruption cases. Journal of Comparative Economics 36, 343-364.

[9] Fernandes, N., and Ferreira, M. A., 2009. Insider trading laws and stock price informativeness. Review of Financial Studies 22, 1845–1887.

[10] Fidrmuc, J.P., and Jacob, M., 2010. Culture, agency costs, and dividends.

Journal of Comparative Economics 38, 321-339.

[11] Fisman, R., 2001. Estimating the Value of Political Connections. The

American Economic Review 91, 1095-1102.

[12] Graham, A., and Dennis N., 2010. The impact of low cost airline operations to Malta. Journal of Air Transport Management 16, 127-136.

[13] Gray, S. J., 1988. Towards a theory of cultural influence on the development of accounting systems internationally. ABACUS 24, 1-15.

[14] Gul, F. A., Kim, J. B., and Qiu, A. A., 2010. Ownership concentration, foreign shareholding, audit quality, and stock pricesynchronicity: Evidence from China.

(27)

[15] Helmuth, J.A., and Robin, A.J., 1998. Trading volume and firm-specific announcements: Implications for the market model. Review of Financial

Economics 7, 183-195.

[16] Hofstede, G. H., 1980. Culture’s consequences: International differences in work-related values. Beverly Hills, Cal., and London: Sage.

[17] Hofstede, G. H., 1984. Culture dimensions in management and planning. Asia

Pacific Journal of Management, 81-99.

[18] Huberman, G., 2001. Familiarity Breeds Investment. The Review of Financial

Studies, 14(3), 659–680.

[19] Jensen, M., and Meckling, W., 1976. Theory of the firm: managerial behavior, agency costs, and ownership structure. Journal of Financial Economics 3, 305– 360.

[20] Jin, L., and Myers, S., 2006. R2 around the world: New theory and new tests.

[21] Jones, C.M., 2003, Editor, Shorting restrictions, liquidity, and returns. Working paper, Columbia University.

[22] Kim, J.B., Chung, R., and Firth, M., 2003, Auditor conservatism, asymmetric monitoring and earnings management, Contemporary Accounting Research 20, 325–359.

[23] Kim, J.B., and Shi, H., 2009, International financial reporting standards, analyst following, institutional infrastructure, and stock price synchronicity. Unpublished working paper. City University of Hong Kong and Fudan University.

[24] La Porta, R., Lopez-de-Silanes, F., Shleifer, A., and Vishny, R.W., 1996, Law and Finance, The Journal of Political Economy 106, 1113-1155.

[25] La Porta, R., Lopez-de-Silanes, F., Shleifer, A., and Vishny, R.W., 1997, Investor protection and corporate valuation, Journal of Finance 57, 1147-1170.

[26] Lesmond, D., 2005, Liquidity of emerging markets. Journal of Financial

Economics 77, 411-452.

[27] Liu, W.M., 2006. A liquidity-augmented capital asset pricing model. Journal

of Financial Economics 82, 631-671.

[28] Lloyd, W. P., 1975. International portfolio diversification of real assets: An inquiry. Journal of Business Research 3, 113-120.

(28)

[29] Morck, R., Yeung, B., and Yu, W., 2000. The information contend of stock markets: why do emerging markets have synchronous stock price movements?

[30] Mukherjee, T.K., and Naka A., 1995. Dynamic relations between macroeconomic variables and the Japanese stock market: an application of a vector error correction model. Journal of Financial Research 18, 223–237.

[31] Parigi, B.M., and Pelizzon, L., 2008. Diversification and ownership concentration. Journal of Banking & Finance 32, 1743-1753.

[32] Piotroski, J.D., and Roulstone, D.T., 2004. The influence of analysts, institutional investors, and insiders on the incorporation of market, industry and firm-specific information into share prices. The Accounting Review 79, 1119– 1151.

[33] Roll, R., 1988. R2. Journal of Finance 43, 541-566.

[34] Salter, S., and Niswander, F., 1995. Cultural influence on the development of accounting systems internationally: A test of Gray’s (1988) theory. Journal of

International Business Studies, 379-397.

[35] Wongbangpo, P., and Sharma, S.C., 2002. Stock market and macroeconomic fundamental dynamic interactions: ASEAN-5 countries. Journal of Asian

Economics 13, 27-51.

(29)

Appendix

Figure 1: Stock price synchronicity including all the variables

Dependent Variable: AD_SYN Method: Panel Least Squares Sample: 2005 2009

Periods included: 5 Cross-sections included: 40

Total panel (balanced) observations: 200

Variable Coefficient Std. Error t-Statistic Prob.

PD 0.000 0.002 -0.086 0.931 MAS 0.000 0.002 0.261 0.794 GOV -0.031 0.008 -3.986 0.000 GDP_P -0.015 0.054 -0.275 0.784 GDP 0.073 0.039 1.880 0.062 GEO 0.012 0.018 0.655 0.513 HERF 0.397 0.167 2.383 0.018 ZR -0.835 0.244 -3.426 0.001 C -0.403 0.602 -0.669 0.504

R-squared 0.297 Mean dependent var -0.770

Adjusted R-squared 0.267 S.D. dependent var 0.438 S.E. of regression 0.375 Akaike info criterion 0.921 Sum squared resid 26.157 Schwarz criterion 1.072 Log likelihood -80.829 Hannan-Quinn criter. 0.982 F-statistic 9.811 Durbin-Watson stat 0.575

Prob(F-statistic) 0.000

(30)

Figure 2: Stock price synchronicity and macro-economic fundamentals

GDP_P -0.150 0.044 -3.395 0.001 GDP 0.074 0.035 2.090 0.038 GEO 0.010 0.017 0.561 0.575 HERF 0.563 0.175 3.212 0.002 ZR -0.500 0.242 -2.069 0.040 C 0.180 0.541 0.334 0.739

Adjusted R-squared 0.147 S.D. dependent var 0.437 S.E. of regression 0.403 Akaike info criterion 1.051 Sum squared resid 31.557 Schwarz criterion 1.150 Log likelihood -99.137 Hannan-Quinn criter. 1.091

F-statistic 7.867 Durbin-Watson stat 0.485

This table shows the results of my panel regression using macro-economic fundamentals as the independent variable from 2005 to 2009. Refer the definitions of the rest variables to figure1 in the appendix.

Figure 3: Stock price synchronicity including GOV

GOV -0.032 0.006 -4.946 0.000 GDP_P -0.013 0.053 -0.245 0.807 GDP 0.079 0.033 2.399 0.017 GEO 0.010 0.016 0.607 0.545 HERF 0.393 0.165 2.383 0.018 ZR -0.827 0.241 -3.435 0.001 C -0.399 0.539 -0.740 0.460

Adjusted R-squared 0.274 S.D. dependent var 0.438 S.E. of regression 0.373 Akaike info criterion 0.901 Sum squared resid 26.168 Schwarz criterion 1.019 Log likelihood -80.867 Hannan-Quinn criter. 0.949 F-statistic 13.204 Durbin-Watson stat 0.573 Prob(F-statistic) 0.000

(31)

Figure 4: Stock price synchronicity including PD

PD 0.000 0.002 -0.073 0.942 GOV -0.032 0.007 -4.490 0.000 GDP_P -0.014 0.054 -0.253 0.801 GDP 0.079 0.033 2.393 0.018 GEO 0.010 0.016 0.604 0.546 HERF 0.392 0.165 2.374 0.019 ZR -0.828 0.242 -3.426 0.001 C -0.381 0.595 -0.641 0.523 PD 0.000 0.002 -0.073 0.942

This panel regression adds PD (Power Distance) in the independent variables. Refer the definitions of the rest variables to figure1 in the appendix.

Figure 5: Stock price synchronicity including MAS

PD 0.000 0.002 -0.073 0.942 GOV -0.032 0.007 -4.490 0.000 GDP_P -0.014 0.054 -0.253 0.801 GDP 0.079 0.033 2.393 0.018 GEO 0.010 0.016 0.604 0.546 HERF 0.392 0.165 2.374 0.019 ZR -0.828 0.242 -3.426 0.001 C -0.381 0.595 -0.641 0.523 PD 0.000 0.002 -0.073 0.942

(32)

Figure 6: Significant variables in the individual regressions

Dependent

Variable: AD_SYN GDP/p GOV PD GEO

Coefficient 0.059 -0.030 (Probability) (0.243) (0.000) Coefficient -0.044 0.004 (Probability) (0.309) (0.034) Coefficient -0.062 0.028 (Probability) (0.122) (0.057)

This table shows the significant variables in the equation (6), ₌ _α ₊ _β ₊ _β ₊ _ε

j j

j GDP p IND

S* 1 * / 2* . In

Culture and stock price synchronicity in emerging countries Daming Su University of Groningen Faculty of Economy and Business Master of Science, Business and Administration of Finance