Stock Market Historic Volatility

National Culture and Global Stock Market Volatility Dante M. Pirouz Doctoral Student The Paul Merage School of Business University of California, Irvine Correspondence: Dante M. Pirouz, The Paul Merage School of Business, University of California, Irvine, CA 92697, USA Tel: (949) 226-8397; E-mail: dpirouz04@merage.uci.edu Running Title: Culture and Stock Market Volatility 1 ABSTRACT This paper examines the impact of culture on an important financial phenomenon, stock market volatility. Using partial least squares analysis, this empirical study investigates the relationship between cultural dimensions and global stock market volatility. The results of this study provide evidence that national cultural dimensions can predict stock market volatility. This study has important implications for understanding how culture at the national level can influence financial market behavior. Keywords: international marketing, financial market behavior, investor behavior, stock market, volatility, consumer behavior 2 INTRODUCTION Volatility is technically defined as the degree to which a market rises or falls in a shortperiod of time (Mullins, 2000). Since the 1970’s volatility in the bond and stock markets has increased globally and stock market volatility is not only detrimental to investors but also can be harmful to the stability of national and global economic systems (Gerlach, Ramaswamy, & Scatigna, 2006). The Asian financial crisis of 1997 is only one example of the negative effect stock market volatility can have on the global financial network and as a result there is strong interest in both the private and public sectors to understand the antecedents of global stock market volatility. While volatility worldwide is on the rise, some countries’ stock markets are more volatile than others. It remains unclear the underlying factors that cause some global stock markets to have differential volatility. A great deal of literature is devoted to the study of volatility especially in understanding what gives rise to volatility, how it can be predicted and measured but there is still no clear understanding of why global stock markets suffer from volatility, why global stock markets vary in their volatility or how to predict which markets will be more volatile than others. This paper investigates whether national cultural dimensions have an impact on stock market behavior, such as volatility. In other words, do the differences in cultural orientation between countries offer any insight into how and why financial markets react and adjust to pricing and information changes? There are three ways to look at this question. The first is to say that culture has no impact on financial market behavior. Many economists and traditional finance academicians would argue in favor of the Efficient Market Hypothesis (EMH), which is the cornerstone of modern financial theory and states that share prices always reflect and incorporate all relevant market information. Culture, along with other sociological and psychological 3 phenomena, has been avoided as a possible explanation for stock pricing and financial market movement. Another way to address the question would be to argue that national culture has a direct effect on financial behavior. A number of hypotheses have been offered as a possible explanation of this proposed relationship. An example is the Cushion Hypothesis (Weber & Hsee, 1998) which says that those from collectivist societies may be more apt to engage in financial risk because they have a societal “cushion” that will protect them from the downside risk. A third example is the Herding Hypothesis, which says that some people, assumedly those from collectivist societies, prefer to conform to the crowd when making financial decisions. However, these explanations have not been reconciled. Finally, there is a third view that culture has an indirect impact on financial market behavior (Chui, Lloyd, & Kwok, 2002, Kwok, 2006). In order to argue for this view, the challenge becomes to find the mediating variable through which culture acts upon market behavior. This paper seeks to challenge the first view, that culture has no relationship with stock market behavior and to test the second and third views. Using cross-country data, the studies in this paper will demonstrate that national culture is indeed related to financial market behavior such as stock market volatility. The market phenomenon, stock market volatility across global stock markets, will serve as the dependent measure for this analysis and will offer a rich data source for testing the proposed relationship. Finally, this paper will attempt to build a link between existing finance literature and cross-cultural literature in order to investigate an important global market phenomenon, stock market volatility. This paper builds on the work of Kwok and Chui, which found evidence for a relationship between national cultural dimensions and financial variables such as debt and capital structure, financial and banking systems and investing strategies, Tellis et al. who found a relationship between cultural dimensions and new product adoption and Stulz and Williamson 4 who found a connection between cultural differences in religion and language in explaining why investor protections vary across countries (Chui, Lloyd, & Kwok, 2002, Chui, Titman, & Wei, 2005, Kwok, 2006, Stulz & Williamson, 2002, Tellis, Stremersch, & Yin, 2003). Collectively, these studies support this paper’s assertion that culture, as operationalized by Hofstede (Hofstede, 1984, Hofstede, 2001, Hofstede, 1991, Hofstede, 2005), can be used as a measure of national level cultural differences. This paper’s results finds that cultural dimensions can be used to predict stock market volatility better than macroeconomic indicators and that culture affects volatility through the degree to which a country has strong globalization ties with other countries and markets. LITERATURE REVIEW Explaining Stock Market Behavior The efficient market hypothesis (EMH) is commonly used to explain and predict the movement in stock market pricing and is used to justify the use of probability calculus in analyzing capital markets (Peters, 1991). An inherent assumption of EMH is that investors are rational. Efficient markets pricing is based on public information that is already discounted. Equilibrium pricing is found by the collective whole assimilating and assessing information and risks. There are three common forms of EMH: the weak form efficiency, semi-strong form efficiency and strong form efficiency. If equity markets are weak form efficient, then we should anticipate that we should not be able to forecast future returns using information available today (Fama, 1970). Yet interestingly, there is evidence that international markets may violate the weak form efficiency. Several studies have demonstrated evidence that international equity returns can be forecasted although these focus mostly on spillover effects which are defined as the effect of one stock market’s volatility on other global markets (Durand, Koh, & Watson, 5 2001, Eun & Shim, 1989, Ghosh, Saidi, & Johnson, 1999, Kahya, 1997, Theodossiou, Kayha, Koutmos, & Christofi, 1997, Theodossiou & Lee, 1993). However, it does open the door for considering factors beyond the traditional macroeconomic ones that might explain international equity returns. Behavioral Finance Traditional finance models based in economics have had been criticized for their mixed descriptive and predictive power when compared with empirical findings (Olsen, 1998). This suggests that financial asset pricing may be due to statistically complex and non-linear effects. While behavioral finance is a relatively new, researchers in this area have sought to challenge the fundamental assumptions of economics and finance to offer a more empirically complete view of financial behavior. These studies have been able to show that powerful psychological effects often give results that counter the predictions of efficient market hypothesis (EMH) (De Bondt & Marcus, 1999, Shefrin & Statman, 1993, Shiller, 1993). For example, the behavioral finance framework suggests that stock prices may be affected by the collective effect of individual investors’ decision making biases and heuristics. This may cause stock prices to be over or under priced or increase market volume or volatility (Morrin, Jacoby, Johar, He, Kuss, & Mazursky, 2002). An example of this effect is the momentum effect – or herding – which is reflected in positive serial correlations in stock prices.1 From a social psychology perspective, this effect shows that a herdlike mentality may influence stock valuations (Shefrin & Statman, 1993, Shiller, 1993). EMH argues that market psychology does not play a role in the determination of stock market volatility (Duffee, 1990). For example, Duffee (1990) developed a model that 1 Positive serial correlation is defined as the persistence of correlation for successive values over time. It is also referred to as autocorrelation. 6 demonstrated how a change in the beliefs of rational agents can lead to a shift in the volatility of the stock market. Yet, some researchers such as Shiller believe that market psychology plays a pivotal role (Shiller, 1988). In fact, a report issued by the Brady Commission (1988) on U.S. stock market conditions found that a good deal of volatility could be attributed to investor psychology such as a loss in financial confidence and advocated the institution of “circuit breakers,” such as closing the market in the event of large point declines or advances, to protect the viability of the financial system. The fact that econometric fundamentals have had mixed results in explaining volatility over time is one prime reason to consider market psychology as a possible determinant of stock market volatility. Stock Market Volatility There are actually a number of definitions of volatility that have been used in the literature. For financial markets, it often refers to the spread of asset returns (Poon, 2005). It also means the spread of all likely outcomes of an uncertain variable. Volatility is not the same as risk, which is associated with an undesirable outcome. Volatility is simply a measure for uncertainty which may be associated with a positive or negative outcome. Financial market volatility has important implications for regulation, monetary policy and macroeconomics (Poon, 2005) (see Appendix E for a further explanation of volatility calculations). Financial markets are affected not only by rational fundamentals such as asset pricing, but they can also be the result of the behavior of participants which can deviate from what might be considered rational. Thus, the area of behavioral finance and behavioral economics has been energized by a number of theorists such as Shiller (1993), De Bondt and Thaler (1985) and Shefrin and Statman (1993), who have begun to provide alternative explanations for market anomalies such as bubbles and volatility, in the hope of understanding phenomena that the 7 predominant financial and economic theories fail to fully explain (De Bondt & Thaler, 1985, Shefrin & Statman, 1993, Shefrin & Statman, 1993, Shiller, 1993). Surprisingly little is known about how stock market investors process information and use decision making strategies. One paper from the consumer behavior literature investigated the investor strategies of professional traders and found that momentum and contrarian investors vary on a number of variables such as price expectations, age, experience, raw performance, risk propensity, cognitive style, knowledge calibration and strategy adaptivity (Morrin, Jacoby, Johar, He, Kuss, & Mazursky, 2002). In addition, consumer confidence is often positively correlated with stock prices (Jansen & Nahuis, 2003). In a study of 11 European countries between 1986 and 2001, it was found that consumer confidence is driven by expectations about the future and has a positive relationship with stock prices. This may be related to “illusion of control.” Due to the phenomenon of “illusion of control,” people erroneously believe that chance events are subject to their personal control (Fenton-O'Creevy, Nicholson, Soane, & Willman, 2003, Langer, 1975). When skill cues, such as choice, competition, familiarity with the stimulus and involvement in decisions are present, people are more likely to behave as if they could exert control over chance situations. Illusion of control has been examined as a factor contributing to trading performance as well as inversely related to trader performance (Fenton-O'Creevy, Nicholson, Soane, & Willman, 2003). Current models of market psychology and stock market psychology usually focus on the coordination problem of trading. Duffee’s model examines how the market psychology of investors is concerned not so much with current volatility but with anticipating future volatility (Duffee, 1990). Thus, variations in market psychology over time can lead to changes in aggregate stock return volatility. Investor confidence is defined as the expectation of future stock 8 market stability which Duffee believes is an important factor in determining stock market volatility. For example, empirical studies have shown only a weak relationship between volatility in the stock market and volatility in other economic indicators (Singleton, 1989). Another study found that volatility increased after financial panic – assumedly when investor confidence was also low (Schwert, 1989). Duffee looked at the effect of confidence in terms of the investors’ expectations of future stock market stability, but instead of looking at when they buy, he focused on their expectations when they sell. His study showed how a change in the beliefs of rational agents may lead to a change in the volatility of a stock market. For example, if investors believe that future volatility will be high, this tends to influence current volatility to be high. If they believe that future volatility will be low, this tends to make current volatility low. Schwert has found that stock market volatility increased during periods of recession, bank panics and other financial crises from 1834 to 1987 (Schwert, 1989). To counteract market volatility, regulators have implemented a number of market stabilization mechanisms including margin requirements and circuit breakers, or trading halts (Duffee, 1990). There is also evidence of increased trading when volatility is high, but it is unclear whether large volume causes high volatility or whether large volatility and trading volume are caused by other factors (Smith, 1990). Behavioral finance has suggested that the collective effect of individual decision makers’ psychological biases – including those of individual investors, government regulators, institutional managers and professional investors – may work to temporarily over or under-price stocks relative to their true value (Morrin, Jacoby, Venkataramani Johar, He, Kuss, & Mazursky, 2002). The momentum effect – or herding behavior – is where stock valuation fluctuates away from fundamentally based pricing due to investors choosing to follow the crowd when buying or 9 selling stock holdings. In contrast, contrarian investors tend to buy out of favor stocks and sell popular stocks. Statman (1999) finds that EMH may be true in that there may be no way for individuals to beat the market; however, asset pricing seems to be vulnerable to individual decision-making psychology for which behavioral finance may offer an explanation (Morrin, Jacoby, Venkataramani Johar, He, Kuss, & Mazursky, 2002). As a result of the work by Shiller (1981) and LeRoy and Porter (1981) on excess stock volatility, there is a body of empirical research that indicates that macroeconomic factors and stock fundamentals do not sufficiently explain stock market volatility. In contrast, there is accumulating evidence that “irrational” behavior from investors, both individual and institutional, novice and professional, can significantly affect stock market returns and volatility (Chang & Dong, 2006). National Cultural Dimensions as an Explanation of Stock Market Volatility Why should one believe that any information regarding stock market phenomena could be gleaned from cultural dimensions of the market’s home nation? One could argue that with global markets, international stock markets are actually a reflection of multinational investors. However, there exists an unusual bias for home assets, a phenomenon in itself that has puzzled macroeconomists for the last 25 years. The home bias portfolio puzzle has been shown in studies by French and Poterba (1991) where it was found that Americans held 94% of their equity wealth in the U.S. stock market and the Japanese held approximately 98% of the equity wealth in their home country (French & Poterba, 1991, French & Poterba, 1991). Strong evidence for cultural influence on stock market pricing can also be found in Jegadeesh and Titman’s Momentum Strategy which proposes that above average returns can be made by investing with the momentum of the market (Jegadeesh & Titman, 2001, Jegadeesh & 10 Titman, 1993). This strategy goes against the traditional contrarian view of buy low and sell high and is based on Jegadeesh and Titman’s findings from U.S. stock market data that investors are subject to overconfidence which causes them to weight and overreact to private information and self attribution bias, which causes them to believe success is due to their own abilities and failure is due to outside circumstances. Interestingly, the above average returns from momentum strategy investing does not hold for non-U.S. stock markets (Chui, Titman, & Wei, 2005, Rouwenhorst, 1998). This might indicate that a difference in cultural effects on information processing might be responsible for this differential return on the strategy. In order to capture national culture, Hofstede’s cultural dimensions were used in this study. While Hofstede has developed 5 dimensions (Individualism/Collectivism, Uncertainty Avoidance, Power Distance, Masculinity/Femininity, Long Term Orientation), this paper will focus on two of the dimensions: Collectivism/Individualism (IDV) and Power Distance (PDI) because these indices are available for the highest number of countries and are believed to be the most applicable to the study’s hypotheses. Hofstede’s culture indexes have been found to be very stable over time and have been replicated in numerous studies (Hofstede, 1984, Kwok, 2006). The cultural dimension of individualism and collectivism captures the relationship between the individual and the collective society. It manifests in how people choose to live together within the family, community or tribal unit and – according to Hofstede – has a significant impact on values, social norms and beliefs. Collectivist countries tend to be more group focused, whereas individualistic societies tend to value the individualist self concept. Power distance has to do with how a culture views inequality. High power distance societies tend to have a large degree of inequality in power and wealth and may even follow a caste system. Societies with low power 11 distance tend to have much more upward mobility and there is a de-emphasis on differences in power and wealth (Hofstede 1984). It has been found in numerous studies that culture has an effect on various kinds of behavior (Fan & Xiao, 2003, Graham, 1984, Hofstede, 1991, Hsee & Weber, 1999, Runtian & Graham, 2006, Weber & Hsee, 1998). However, it is not entirely clear how and when culture impacts behavior, especially in terms of financial decision making. If culture does affect financial decision making, perceptions of risk and the pricing of assets, then it might be reasonable to look at phenomenon such as volatility to understand if there is a relationship. There are several hypotheses that might explain how culture can influence global financial market behavior such as volatility. This paper will investigate whether these hypotheses might offer an explanation for global stock market volatility. THEORETICAL FOUNDATION AND HYPOTHESES Cushion Hypothesis Hsee found Americans were considerably more risk-averse than Chinese (Hsee, 1998). As possible explanations, he offers the cushion hypothesis and the opportunity hypothesis. The cushion hypothesis finds that in individualistic cultures, with an emphasis on personal freedom, there is less of a societal safety net to protect individuals against adverse results of a risky choice. On the other hand, collectivist societies emphasize social relatedness and interdependence of family and community which allows individuals to have different perceptions of risk. Individuals in collectivist societies are more likely to receive help from family and the community in the case of failure. Thus they appear more risk seeking due to this societal “cushion.” Herding Hypothesis 12 There is evidence that some investors are more inclined to contrarian market behavior and some are more inclined to herding behavior (Morrin, Jacoby, Johar, He, Kuss, & Mazursky, 2002). Contrarian behavior is characterized by buying when stock prices are declining and selling when stock prices are escalating. Herding behavior, evident in positive serial correlations in stock prices and known as the momentum effect, is often a factor in market crashes and bubbles (Jegadeesh & Titman, 1993). Social interaction effects such as herding or conforming to the crowd are one of the factors found to affect financial decision making (Hirshleifer, 2001). The first hypothesis is based on a direct relationship between cultural dimensions underlying Relationship Orientation and Stock Market Volatility. H1: Countries that tend to be Relationship Oriented (more collectivist and with high power distance) will have higher Stock Market Volatility. This hypothesis assumes a direct relationship between culture and volatility and is supported by the cushion hypothesis and the herding hypothesis. Global Capital Market Argument This hypothesis is supported by the global capital market argument which argues that there are risk-sharing benefits to maintaining strong global ties between nations (Stulz, 1999, Stulz & Williamson, 2002). In addition to reducing market risk premiums, globalization affects systematic risk (or beta) of individual companies. It could be argued that globalization also affects systematic risk for countries as well. The second hypothesis is based on a mediated relationship between the cultural dimensions underlying Relationship Orientation and Stock Market Volatility via the Globalization variable. 13 H2: Countries that tend to be relationship oriented (more collectivist and with high power distance) will have lower globalization ties. Support for this hypothesis comes from evidence developed in Cateora and Graham (2004) which finds that relationship oriented cultures tend to be more insular and change averse, which can lead to more global isolation (Cateora & Graham, 2004). H3: Countries with lower globalization ties will have higher stock market volatility. In addition, the Global Capital Market Argument states that global ties cause more stability in financial systems. The purpose of this paper is to test these hypotheses to determine if there is evidence for an association between culture and financial market behavior, and to reveal if there are any mediating factors that work between cultural dimensions and volatility. RESEARCH DESIGN AND METHOD Data Collection Stock market volatility has been estimated using the Morgan Stanley Capital International Equity Indices which are widely used international equity benchmarks used in a number of studies (Jorion & Goetzmann, 1999). The standard national indices for each country in the study were used with the standard index performance price at the last day of each month for each year in U.S. dollars. While the MSCI is an indirect measure of global stock markets, it has been used in a number of studies as a means of tracking stock market value across many markets. The MSCI index data for the major stock market in each country were calculated using the equation for historical volatility using the month to month price in U.S. dollars. The stock price data used for this study are taken from MSCI, which has been used in a number of other 14 studies by economists and finance researchers (Dwyer & Hafer, 1988). Stock market volatility was calculated with the start and end of year index prices in U.S. dollars. Historical volatility is a measure of price changes of a security or return over a specific period of time using the standard deviation of the continuously compounded return. There are a number of ways to calculate single-state historical volatility models, including the random walk model, historical average method, moving average method, exponential smoothing method, exponentially weighted moving average method (EWMA) and simple regression method (Poon, 2005). Volatility can be examined in short or long term time framed and with differing price intervals. However, as long as price changes are measured in regular intervals, the annualized volatilities calculated using these differing parameters are usually comparable. The following formula calculates the historical volatility for a given period over a specific time span. HV = SSS * TP N −1 where SSS = ∑ (X − X ) 2 N = number of periods for time span TP = total number of trading periods for the year. In this study, historical volatility is used, which relies on standard deviation which is similar to other studies (Schwert, 1997). A total of 50 countries were selected for this analysis with the criteria that they each have at least one stock market established and an index on that market is maintained by Morgan Stanley Capital International (MSCI) (Morgan Stanley Capital International, 2005). 15 (Insert Table 1) The stock market data used were for the years 2003, 2004 and 2005. The index prices were calculated in U.S. dollars. In addition, the index for Individualism/Collectivism (IND) and Power Distance (PDI) from Hofstede (2001) were used as the culture indicators for the variable Relationship Orientation for the 50 countries. The Globalization Index is generated annually by A.T. Kearney and is made up of 14 variables grouped into 4 components: economic integration (GlobEcon), technological connectivity (GlobTech), personal contact (GlobPers), and political engagement (GlobPolit) to determine the ranking of 62 countries (A.T. Kearney, 2005). Included are variables such as trade and financial flows, movement of people across borders, international telephone traffic, internet usage, and participation in international treaties and peacekeeping operations. Economic integration includes data on trade, foreign direct investment (FDI), portfolio capital flows and investment income payments and receipts. The personal contact component is made up of data tracking international travel and tourism, international telephone traffic, cross border remittances, and personal transfers. The technological connectivity component is made up of data on the number of internet users, internet hosts, and secure servers. The political engagement component is made up of each country’s membership in international organizations, personnel and financial contributions to U.N. Security Council missions, ratification of selected multilateral international treaties, and the amount of governmental transfer payments and receipts. It is worthy of note that there is little correlation between the size of a country’s economy and its level of globalization. Large economies do not necessarily have an advantage in globalization. 16 However, small countries do tend to have an advantage over larger countries with similar levels of per capita income. In this study we are using the Globalization Index for 2003, 2004 and 2005. Research Method The hypotheses were tested using a partial least squares analysis with SmartPLS 2.0 (beta) to examine the relationship between cultural dimensions, as is defined by the Holistic View Hypothesis and global stock market volatility (Nisbett, 2003, Ringle, Wende, & Will, 2005). Partial least squares was chosen as the method of analysis since it can be used with data that come from non-normal distributions and less than interval level data (Falk & Miller, 1992, West & Graham, 2004). Unlike LISREL and other covariance structure analysis modeling approaches, partial least squares seeks the minimization of error or equivalently the maximization of variance explained which can be determined by examining the R2 values of the dependent or endogenous constructs (Falk & Miller, 1992, Hulland, 1999). This functions as an indicator of the model’s goodness of fit. Partial least squares was chosen over regression or structural equation modeling for this study due to the non-normal aspects of the data and the relatively small sample size of greater than 50 (Falk & Miller, 1992, West & Graham, 2004). Partial least squares is well suited to handle problems of non-normality, multicollinearity and non-linearity of the data, allow for the inclusion of a third variable in the model and allow the measurement and theoretical model to be estimated simultaneously. In addition, partial least squares was found to be more robust than other methods of estimation which makes it a suitable choice for this study (Naik & Tsai, 2000). The variables used in this analysis are Individualism/Collectivism Index (IDV) and Power Distance Index (PDI), which together form the latent variable called Relationship Orientation. It is posited that countries with higher Relationship Orientation will be more 17 collectivist and have a higher degree of power distance than countries with low Relationship Orientation. Historical stock market volatility was calculated for the years 2003, 2004, and 2005. The higher the number is, the higher the historical volatility in that year for the given country. The Globalization variable is from the A.T. Kearney Globalization Index for 2003, 2004 and 2005 and is an index ranking the countries highest in globalization ties with the highest rank (lowest number) and those with the lowest degree of globalization ties with the lowest rank (highest number) (A.T. Kearney, 2005). The variable Macroeconomics was formed by GDP, population and market capitalization for 2004 and is included as a confounding variable to show that these factors are not important determinants of stock market volatility. PARTIAL LEAST SQUARES ANALYSIS OF CULTURE AND STOCK MARKET VOLATILITY In order to test the hypotheses, a model was constructed to test the relationship between each cultural dimension: Collectivism/Individualism, Power Distance, and Stock Market Volatility. This study focuses on the question of whether there is a mediator that functions between the cultural dimensions and Stock Market Volatility and better illustrates the relationship. It is hypothesized that the level of Globalization, measured in the A.T. Kearney/Foreign Policy annual Globalization Index, might be affected by cultural dimensions unique to each country and mediate the relationship to Stock Market Volatility (A.T. Kearney, 2005). This study tests hypothesis 1 and 2, which are based on hypotheses from the existing literature: the Cushion Hypothesis and the Herding Hypothesis, and hypothesis 3, which is 18 supported by the Global Capital Market argument. Hsee et al. proposed the Cushion Hypothesis which states that those in socially collectivist cultures are able to take on more financial risk, because social networks can serve to insulate them from the negative effects (Weber and Hsee 2000). If a culture has a high Collectivism and Power Distance rating, it would be expected that there would be higher speculative activity and higher volatility due to the belief of these investors that the community will provide a financial “cushion” in the case of stock market losses. With the Herding Hypothesis, it would be assumed that high Power Distance would be an indicator of a culture that tends to be authoritarian, conformist and traditional. Individual investors would be more likely to follow where the market is moving which during periods of uncertainty could lead to high volatility. Previous studies have shown that even institutional investors can fall victim to herding behavior which can significantly affect stock market pricing (Chang & Dong, 2006). Both of these hypotheses would point to a direct relationship between Relationship Orientation, formed by the Individualist/Collectivist and Power Distance indices, and Stock Market Volatility. Hypothesis 3 is supported by the global capital market argument. This argument would support a strong relationship between Globalization and Stock Market Volatility. It would predict that countries with strong global ties would be more likely to enjoy dampening effects on Stock Market Volatility due to the benefits to risk mitigation through diversification (Stulz, 1999, Stulz & Williamson, 2002). RESULTS The partial least squares model supports hypotheses 2 and 3 which state that collectivist and power distant countries are more likely to have lower global integration and with lower 19 globalization these countries have higher historical stock market volatility. In addition, the model shows that the Macroeconomic variable, which is made up of GDP, population and market capitalization, do not contribute to explaining stock market volatility. Hypothesis 1 is not supported by the model which demonstrates that the relationship between Relationship Orientation of a country does not necessarily have a direct impact on stock market volatility. The nomogram for the model is as follows: (Insert Figure 1) Globalization and Stock Market Volatility were formed using reflective indicators made up of underlying factors (Fornell & Bookstein, 1982). The variables Relationship Orientation and Macroeconomics are formative indicators since these constructors are conceived as explanatory combinations of indicators. The first result indicated by this model, is that macroeconomic factors do not explain Stock Market Volatility variance. If the traditional economic view of financial market behavior is true, then the Macroeconomic variable, which is formed by variables of the size of the country (population), the size of the economy (GDP) and the size of the stock market (market capitalization), should have a significant and strong impact on Stock Market Volatility. However, this model shows that the Macroeconomics variable path to Stock Market Volatility is not significant and it does not increase the variance explained in the Volatility variable. Secondly, the R2 tells us that 42% of the variance in Historical Stock Market Volatility is explained by the model which is a significant improvement over the model in the first study. The R2 for Globalization is 31%. Both of these explain a high degree of variance in the model. 20 (Insert Table 2) T values for the path coefficients indicate that the paths between Relationship Orientation and Globalization and Globalization and Volatility are significant as hypothesized in hypothesis 1 and 2 as shown in table 4. In addition, the path between Macroeconomics and Volatility is not significant indicating that these macroeconomic indicators do not explain Volatility well. (Insert Table 3) As anticipated, the direct relationship between Relationship Orientation and Volatility does not have a significant T value which argues against a direct relationship. In addition, the relationship between Macroeconomics and Volatility also does not have a significant relationship. In order to ensure adequate convergent validity, we can look at AVE or average variance extracted, which is the average variance shared between a construct and its measures and should be greater than the variance shared between the construct and the other constructs in the model (i.e. the squared correlation between two constructs) (Hulland, 1999, Ping, 2005). The formula for AVE is denoted by: AVE = ∑ λ yi / ∑ λ yi + ∑ var(∈i ) 2 2 ( ) It has been suggested that for convergent validity, AVE should be greater than 0.5. For discriminant validity, AVE should be 0.7 or higher (Ping, 2005). (Insert Table 4) For this model, the AVE for Volatility and Globalization suggests both convergent and discriminant validity. (Insert Table 5) (Insert Table 6) (Insert Table 7) 21 Bootstrapping was performed to determine the significance of the model. 200 samples were run generating a degree of freedom of 199 with a critical T-value of 1.66 for a one tailed test. Overall, the key paths were found significant via the bootstrap method (see Table 8). (Insert Table 8) DISCUSSION The results of this partial least squares analysis show that the effect of national culture on historical stock market volatility is an indirect one. The direct relationship between Relationship Orientation and Stock Market Volatility was neither strong nor significant. On the other hand, the relationship between Relationship Orientation, Globalization and Stock Market Volatility was strong, significant and explained a sizeable amount of the variance in the Volatility variable. Thus, this model offers evidence that national culture strongly affects the degree to which a country is open to globalizing forces and subsequently this level of globalization has a significant effect on whether a stock market is more likely to experience low or moderate overall index volatility or high overall index volatility relative to other countries. Stock market volatility is an important financial dynamic. Perceptions about volatility affect the confidence and behavior of individual investors, institutional investors and regulators. There is a strong desire by both academicians and practitioners to understand the factors that affect volatility in order to maintain fair, efficient and effective financial markets worldwide (Smith, 1990). While a predominance of research focuses on the macro and micro-economic factors affecting stock market volatility, there is a more fundamental behavioral question that deserves investigation: whether the cultural predisposition of a nation has a relationship to the level of that country’s stock market volatility. It is posited in this paper that financial market behavior is influenced by the cross-cultural psychology of its participants. Thus, it is 22 hypothesized that the local national culture in which these markets exist should have a relationship with stock market performance. Under this premise, this study investigates the degree to which national cultural dimensions influence the extremity of volatility in its local stock market. In contrast to theories based on the Efficient Market Hypothesis (EMH), some have advocated in recent years a more controversial view of financial market behavior, namely that psychological and sociological factors can have a large effect on such phenomenon as stock volatility (Shiller, 1984, Summers, 1986). For example, there is evidence that stock volatility is higher on average during recessions as well as during major and minor banking crises which is not explained well by the existing EMH theory (Schwert, 1989). Essentially, these researchers find evidence for the belief that investing in speculative assets is to a great extent a social activity (Shiller, 1993). Thus, it would follow that movements in social behavior, such as differences in thinking, can affect prices in financial markets. However, the bulk of academic research has for the most part avoided behavioral explanations for market behavior. Shiller (1989) argues to the contrary that “…mass psychology may well be the dominant cause of movement in the price of the aggregate stock market (Shiller, 1993).” This is the underlying tenet of the area of behavioral finance. If this belief has validity, then it makes sense to investigate how culture affects financial institutions such as stock markets. This study shows that cultural dimensions are not simply driven by geographical proximity as would be suggested in the Holistic View Hypothesis. Country level variables, when divided into an Eastern and Western dichotomy, do not explain a great deal of the variance in country level financial variables like stock market volatility. However, when cultural dimensions 23 for each country are allowed to vary based on individual country differences, the model shows a better result. In addition, there is evidence from this analysis that it is not appropriate to expect a direct relationship between cultural dimensions. Instead, it is more reasonable to consider that cultural dimensions impact financial variables through mediating variables. This means that culture may indeed offer a salient explanation for financial market behavior but that analysis should include additional factors in order to create an informative model. Above all, this study demonstrates that globalization is an important factor in maintaining financial market stability and that culture has a heavy influence on the degree to which a country initiates and builds global ties. In addition, this study shows that macroeconomic factors, such as GDP and market capitalization, do not give an accurate nor informative view of financial market behavior. Instead, it seems that without the inclusion of cultural variables in any analysis of financial market behavior, the analysis will be incomplete at best. Limitations In terms of the data, the Morgan Stanley Capital International stock market index is a proxy for actual stock market data and offers an approximation of the volatility of any given stock market. It could be seen as a limitation of the study that actual stock prices were not used to calculate the historical volatility. Historical volatility could be calculated using any number of other data, for example a subset of actual stock prices. However, in order to have a consistent data source across all the stock markets included in this study, the MSCI data seem a reasonable approximation to determine if there is any causal relationship between the variables. Also, the ratio of foreign to domestic investors may affect the results of this study. A stock market that has a large percentage of foreign investors may dilute the affect of national cultural dimensions on 24 the market behavior. It may be advisable to weight the variables with the percentage of local versus foreign investors in order to gain a more accurate outcome. There may be an issue with determining causality due to the unmeasured cross country heterogeneity such as savings rates, banking industry structure, regulatory policy and liberalization is a problem (Filer, Hanousek, & Campos, 1999). Also, the time horizon of 20022005 for this study was relatively short given the limitations of availability of stock market index data from the Morgan Stanley data source. A longer time horizon of stock market volatility might yield improved results. In addition, there may be an unusual affect on global stock market behavior due to the events on September 11, 2001 which may have affected the stock market data for 2001 and the subsequent year. Finally, the Hofstede cultural dimension data were collected in 1968 and 1972 (Hofstede, 1984). There may have been significant shifts in the cultural dimensions since that time, possibly impacting the validity of the data. In addition, there has been some criticism of the dimensions as being too simplistic (McSweeney, 2002). Future Studies It might be interesting to incorporate the concept of stock market volatility spillover with an analysis of cultural dimensions. The heat wave hypothesis refers to volatility having only location-specific autocorrelation. For example, a volatile day in New York might be followed by another volatile day in New York only and not spread to other stock markets globally. The meteor shower hypothesis states that intraday volatility can spillover from one market to the next. For example, a volatile day in New York would likely be followed by a volatile day in Tokyo. However, findings have not found spillover affects for every market. For example, Fleming & Lopez (1999) found that the meteor shower hypothesis works well for London and 25 Tokyo but the heat wave hypothesis works well for New York (Fleming & Lopez, 1999). How the national culture in each market might affect these hypotheses might be an interesting investigation. In addition, there are other types of financial market structures that could be applied in a cross cultural study. Previous studies have looked at cross cultural differences in banking structures and firm debt structures using Hofstede’s cultural dimensions as a predictor (Chui, Lloyd, & Kwok, 2002, Kwok, 2006). For example, stock market growth rates, national interest rates, and currency exchange rates could be possible variables that could be examined using national culture variables. CONCLUSION Using cross-country data, this paper offers evidence for the hypothesis that culture has an, albeit indirect but significant, effect on stock market volatility. While to those researchers who study culture it would make sense that these two variables would be interconnected, this is one of the first papers to offer an empirical investigation of global stock market data in order to make the connection between stock market behavior and national culture. Secondly, in the tradition of work done by Kwok and others (Chui, Lloyd, & Kwok, 2002, Kwok, 2006), this paper attempts to bridge the gap between the finance literature and the cross-cultural literature. This paper adds to the literature through its use of partial least squares analysis which establishes and supports a predictive model for the relationship between national culture and stock market volatility. 26 Appendix A: Stock Markets Included There are a total of 110 stock exchanges around the world. Of the stock markets around the world, data for the following 50 stock markets were included in this study: (Insert Table A-1) These stock markets represent the majority of the world stock market capitalization and most of its exchange-traded futures, options, listed investment funds and bonds (Dimson, Marsh, & Staunton, 2002, World Federation of Stock Exchanges, 2006). Appendix B: Data Sources Morgan Stanley Capital International (Morgan Stanley Capital International, 2005) • • • • • • Index Price Annualized Historical Index Returns WDI (World Bank 2004) GDP Population Market Capitalization World Federation of Exchanges (World Federation of Stock Exchanges, 2006) Total Shares Traded Appendix C: Definition of Variables • • • GDP per Capita: GDP divided by population Historical Volatility: Standard deviation of MSCI price on the last trading day of the year Globalization Index: High globalization = low number –Low globalization = high number 27 • Power Distance: High Power Distance = high number - Low Power Distance = low number • Collectivism/Individualism: Individualist = high number – Collectivist = low number Appendix D: MSCI Index Calculation Formulas The MSCI equity indices measure the performance of a set of equity securities over time. The MSCI Indices are calculated using the Laspeyres’ concept of a weighted arithmetic average together with the concept of chain-linking. Price indices measure market price performance only, and are calculated at least on a daily basis. Each index measures the sum of the market capitalization weighted returns of all its constituents on a given day. MSCI national and regional indices are calculated in local currency as well as U.S. dollars. Index levels are also available in several other currencies such as GBP, EUR, JPY, etc. While the local currency series of regional indices cannot be replicated in the real world, it represents the theoretical performance of an index without any impact from foreign exchange fluctuations—a continuously hedged portfolio. The general expression of the indices is set forth below. The previous period’s index level is multiplied by the change in the market performance. Appendix E: Calculating Volatility There are a number of ways to calculate volatility including annualized volatility, historical volatility, implied volatility (Holton, 2006). A formalized definition of volatility is as follows. Let ...t − 2 Q, t −1Q, t Q, t +1Q,... be a stochastic process with Q representing for example prices, accumulated values, exchange rates, etc. The volatility at time t-1 is defined as the 28 standard deviation of the time t return. Volatility is often calculated as the annualized standard deviation of a return series as the following: σ= ∑ (r (t ) − avr ) N 2 where r(t) = returns and avr = average portfolio return = ∑ (r (t )) / N Sometimes, variance σ 2 is used to measure volatility which is the square of the standard deviation (Poon, 2005). Volatility can also be calculated using monthly returns: V = σ * 12 Using log returns, the definition is: ⎛ ⎛ tQ ⎞⎞ V = σ ⎜ log⎜ t −1 ⎟ ⎟ ⎜ ⎜ Q ⎟⎟ ⎠⎠ ⎝ ⎝ where log denotes the natural logarithm This definition usually assumes returns are conditionally homoskedastic. Usually volatilities for different units of time are fundamentally different. For example there is no direct relationship between weekly and annual volatility. One exception to this rule is the square root of time rule. If the fluctuations in a stochastic process from one period to the next are independent, meaning that there is no serial correlation, volatility increases with the square root of the unit of time. The square root of time rule is exact if the volatilities are calculated using the log returns and approximate if the volatilities are based on simple returns. LOGSi = LOG ( Pi / Pi − 1) where LOG is the logarithm function Pi is the current price Pi - 1 is the previous price. 29 Next, calculating the total logarithms for the time span reviewed Tlog s = ∑ LOGS i i =1 n Finally, to calculate the sum of squares of the differences between the individual logarithms for each period and the average logarithm: SSD = ∑ ( LOGS i − ALOGS ) 2 i −1 n where SSD = the sum of the squared differences S = total squares of all the differences LOGSi = the logarithms of the price change for period i ALOGS = the average of the logarithms 30 Figures Figure 1: Partial Least Squares Model with Path Coefficients and R2 31 Tables Table 1: Countries Included in the Study Country 1. Argentina 2. Australia 3. Austria 4. Belgium 5. Brazil 6. Canada 7. Chile 8. China 9. Colombia 10. Czech Republic 11. Denmark 12. Egypt 13. Finland 14. France 15. Germany 16. Greece 17. Hong Kong 18. Hungary 19. India 20. Indonesia 21. Ireland 22. Israel 23. Italy 24. Japan 25. Jordan 26. Malaysia 27. Mexico 28. Morocco 29. Netherlands 30. New Zealand 31. Norway 32. Pakistan 33. Peru 34. Philippines 35. Poland 36. Portugal 37. Russia 38. Singapore 39. South Africa Individualism/Collectivism Index (IND) 46 90 55 75 38 80 23 20 13 58 74 38 63 71 67 35 25 80 48 14 70 54 76 46 38 26 30 38 80 79 69 14 16 32 65 27 n/a 26 65 Power Distance Index (PDI) 49 36 11 65 69 63 63 80 67 35 18 80 33 68 35 60 68 46 77 78 28 13 50 54 80 104 81 80 38 22 31 55 64 94 50 63 n/a 76 49 32 40. South Korea 41. Spain 42. Sri Lanka 43. Sweden 44. Switzerland 45. Taiwan 46. Thailand 47. Turkey 48. United Kingdom 49. United States 50. Venezuela 18 51 n/a 71 68 17 20 37 89 91 12 60 57 n/a 31 34 58 64 66 35 40 81 R2 0.0000 0.4175 0.3095 0.0000 Table 2: Latent Variable Correlation Table with R2 GLOBAL- VOLATILITY IZATION GLOBALIZATION 1.0000 0.0000 VOLATILITY 0.6366 1.0000 MACRO -0.4398 -0.3757 ECONOMICS RELATIONSHIP -0.5564 -0.3866 Table 3: T Values GLOBALIZATION GLOBALIZATION VOLATILITY MACRO ECONOMICS RELATIONSHIP ORIENTATION Table 4: AVE GLOBALIZATION VOLATILITY MACROECONOMICS RELATIONSHIP ORIENTATION Table 5: Composite Reliability GLOBALIZATION VOLATILITY AVE 0.9775 0.4805 0.0000 0.0000 0.0000 0.0000 0.0000 -4.6387 MACRO ECONOMICS 0.0000 0.0000 1.0000 0.3007 RELATIONSHIP 0.0000 0.0000 0.0000 1.0000 VOLATILITY 3.9151 0.0000 -0.9244 -0.2764 MACRO ECONOMICS 0.0000 0.0000 0.0000 0.0000 RELATIONSHIP 0.0000 0.0000 0.0000 0.0000 Composite Reliability 0.9924 0.7016 33 Table 6: Outer Loadings GLOBALIZATION VOLATILITY POWER DISTANCE COLLECTIVISM MRKTCAP04 POP04 GDP04 HISTVOL03 HISTVOL04 HISTVOL05 GLOBTotal03 GLOBTotal04 GLOBTotal05 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.9822 0.9900 0.9939 0.0000 0.0000 0.0000 0.0000 0.0000 0.8888 0.2623 0.7634 0.0000 0.0000 0.0000 MACRO ECONOMICS 0.0000 0.0000 0.6434 -0.6564 0.4975 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 RELATIONSHIP ORIENTATION 0.6983 0.9946 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 Table 7: Outer Weights GLOBALIZATION VOLATILITY POWER DISTANCE COLLECTIVISM MRKTCAP04 POP04 GDP04 HISTVOL03 HISTVOL04 HISTVOL05 GLOBTotal03 GLOBTotal04 GLOBTotal05 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.3293 0.3366 0.3455 0.0000 0.0000 0.0000 0.0000 0.0000 0.6975 -0.0476 0.5143 0.0000 0.0000 0.0000 MACRO ECONOMICS 0.0000 0.0000 1.5213 -0.6569 -0.8240 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 RELATIONSHIP ORIENTATION -0.1624 1.1195 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 Table 8: Bootstrapping Outer Loadings GLOBALIZATION INDEX 2003 --> GLOBALIZATION GLOBALIZATION INDEX 2004 --> GLOBALIZATION GLOBALIZATION 2005 --> GLOBALIZATION POWER DISTANCE --> RELATIONSHIP ORIENTATION COLLECTIVISM --> RELATIONSHIP ORIENTATION HISTORICAL VOLATILITY 2003 --> HISTORICAL STOCK VOLATILITY HISTORICAL VOLATILITY 2004 --> HISTORICAL STOCK VOLATILITY HISTORICAL VOLATILITY 2005 --> HISTORICAL STOCK VOLATILITY T Statistics 231.1230* 436.7544* 898.6410* 5.9528* 26.9030* 31.4006* 1.2066 8.3566* 34 Outer Weights MARKET CAP 2004 --> MACROECONOMICS POPULATION 2004 --> MACROECONOMICS GDP 2004 --> MACROECONOMICS GLOBALIZATION INDEX 2003 --> GLOBALIZATION GLOBALIZATION INDEX 2004 --> GLOBALIZATION GLOBALIZATION 2005 --> GLOBALIZATION POWER DISTANCE --> RELATIONSHIP ORIENTATION COLLECTIVISM --> RELATIONSHIP ORIENTATION HISTORICAL VOLATILITY 2003 --> HISTORICAL STOCK VOLATILITY HISTORICAL VOLATILITY 2004 --> HISTORICAL STOCK VOLATILITY HISTORICAL VOLATILITY 2005 --> HISTORICAL STOCK VOLATILITY MARKET CAP 2004 --> MACROECONOMICS POPULATION 2004 --> MACROECONOMICS GDP 2004 --> MACROECONOMICS GLOBALIZATION --> HISTORICAL STOCK VOLATILITY 4.5714* 3.8281* 2.4311* 54.5945* 73.3635* 73.3689* 0.5789 4.9819* 10.4656* 0.3469 10.5791* 1.7502* 4.1125* 0.9185 5.4732* 1.8860* 6.8472* 0.2569 5.4732* 1.8860* 6.8472* 3.0751* Path Coefficients MACROECONOMICS --> HISTORICAL STOCK VOLATILITY RELATIONSHIP ORIENTATION --> GLOBALIZATION RELATIONSHIP ORIENTATION --> HISTORICAL STOCK VOLATILITY Total Effects GLOBALIZATION --> HISTORICAL STOCK VOLATILITY MACROECONOMICS --> HISTORICAL STOCK VOLATILITY RELATIONSHIP ORIENTATION --> GLOBALIZATION RELATIONSHIP ORIENTATION --> HISTORICAL STOCK VOLATILITY * Found significant via bootstrapping procedure with 200 iterations. Table A-1 Country Full Name Date of First Stock Exchange 1872 1871 1771 1801 1877 1861 1892 1990 1929 1. 2. 3. 4. 5. 6. 7. 8. 9. Argentina Australia Austria Belgium Brazil Canada Chile China Columbia Mercado De Valores S.A. (Merval) Australian Stock Exchange (ASX) Austrian Borse (Wiener Borse) Euronext Brussels Sao Paolo Stock Exchange Toronto Stock Exchange (TSX) Santiago Stock Exchange Shanghai Stock Exchange (SSE) Bogota Stock Exchange (BSE) Market Capitalization 2003 (in U.S. $) $776B $975B $88B $330B $889B $62B $532B $51B # of Companies Listed 1,774 3,600 239 1,287 104 35 10. Czech Republic 11. Denmark 12. Egypt 13. Finland 14. 15. 16. 17. 18. 19. France Germany Greece Hong Kong Hungary India Prague Stock Exchange Copenhagen Stock Exchange Helsinki Stock Exchange (Helsingin Porssi) Euronext Paris Frankfurt Stock Exchange (FWB) Athens Exchange Hong Kong Stock Exchange Budapest Stock Exchange (BSE) National Stock Exchange of India (NSE) Jakarta Stock Exchange (JSE) Irish Stock Exchange (ISEQ) Tel Aviv Stock Exchange (TASE) Borsa Italiana Tokyo Stock Exchange Amman Stock Exchange (ASE) Bursa Malaysia La Bolsa Mexicana De Valores Casablanca Stock Exchange Euronext Amsterdam New Zealand Exchange (NZX) Oslo Bors Karachi Stock Exchange Bolsa De Valores De Lima (BVL) Philippine Stock Exchange (PSE) Warsaw Stock Exchange Euronext Lisbon Russion Trading System (RTS) Singapore Exchange (SGX) Johannesburg Stock Exchange (JSE) Korea Exchange (KRX) Spanish Exchanges (BME) Colombo Stock Exchange (CSE) Stockholm Stock Exchange (OMX) Swiss Stock Exchange (SWX) Taiwan Stock Exchange (TSEC) Stock Exchange of Thailand (SET) Istanbul Stock Exchange (ISE) London Stock Exchange New York Stock Exchange (NYSE) Bolsa de Valores de Caracas 1871 1808 1890 1912 1724 1685 1892 1890 1864 1992 1912 1799 1934 1808 1878 1999 1929 1894 1929 1611 1872 1881 1947 1860 1927 1817 1769 1810 1911 1887 1921 1860 1900 1901 1850 1953 1975 1866 1698 1792 1893 $61B (2006) $118B $170B 33 $1,079B $104B $715B $19B $278B $55B $85B $69B $615B $2,953B $32B (2006) $161B $122B $29B (2006) $2,080B $33B $96B $54B $14B $23B $29B $6B $174B $261B $372B $726B $1.5B $289B $726B $379B $119B $68B $2,460B $11,329B 1,378 20. Indonesia 21. Ireland Israel Italy Japan Jordan Malaysia Mexico Morocco Netherlands New Zealand Norway Pakistan Peru Philippines Poland Portugal Russia Singapore South Africa South Korea Spain Sri Lanka Sweden Switzerland Taiwan Thailand Turkey United Kingdom 49. United States 50. Venezuela 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 54 1,392 663 600 684 242 679 64 36 References A.T. Kearney, Inc. and the Carnegie Endowment for International Peace. 2005. Measuring Globalization: The Global Top 20. Washington, DC: A.T. Kearney, Inc. and the Carnegie Endowment for International Peace. Cateora, Philip R. & John L. Graham. 2004. International Marketing. 12th ed. Burr Ridge, IL: McGraw-Hill. Chang, Eric C. & Sen Dong. 2006. Idiosyncratic volatility, fundamentals, and institutional herding: Evidence from the Japanese stock market. Pacific-Basin Finance Journal, 14(2): 13554. Chui, Andy C.W., Alison E. Lloyd, & Chuck C. Y. Kwok. 2002. Capital Structure: Is National Culture a Missing Piece to the Puzzle? Journal of International Business Studies Chui, Andy C.W., Sheridan Titman, & K.C. John Wei. 2005. Individualism and Momentum around the World. De Bondt, Werner F.M. & Alan J. Marcus. 1999. Does the Stock Market Overreact? Journal of Finance, 40: 793-805. De Bondt, Werner F.M. & Richard Thaler. 1985. Does the Stock Market Overreact? Journal of Finance, 40(July): 793-805. Dimson, Elroy, Paul Marsh, & Mike Staunton. 2002. Triumph of the Optimists: 101 Years of Global Investment Returns. Princeton, NJ: Princeton University Press. Duffee, Gregory R. 1990. The Importance of Market Psychology in the Determination of Stock Market Volatility.In System, Board of Governors of the Federal Reserve, editor: Board of Governors of the Federal Reserve System. Durand, Robert B., S.K. Koh, & I. Watson. 2001. Who Moved Asian-Pacific Stock Markets? A Further Consideration of the Impact of the U.S. and Japan. Australian Journal of Management, 26: 125-45. Dwyer, Jr., Gerald P. & R.W. Hafer. 1988. Do Fundamentals, Bubbles, or Neither Determine Stock Prices? Some International Evidence.In Dwyer, Jr., Gerald P. & R.W. Hafer, editors, The Stock Market: Bubbles, Volatility and Chaos. Boston: Kluwer Academic Publishers. Eun, C. & S. Shim. 1989. International Transmission of Stock Market Movements. Journal of Financial and Quantitative Analysis, 24: 241-56. Falk, R. Frank & Nancy B. Miller. 1992. A Primer for Soft Modeling. Akron: University of Akron. Fama, E.F. 1970. Efficient Capital Markets: A Review of Theory and Empirical Work. Journal of Finance, 25(49): 383-417. Fan, Jessie X. & Jing Jian Xiao. 2003. Cross-Cultural Differences in Risk Tolerance: A Comparison between Chinese and Americans. Consumer Interests Annual(49): 1. Fenton-O'Creevy, Mark, Nigel Nicholson, Emma Soane, & Paul Willman. 2003. Trading on Illusions: Unrealistic perceptions of control and trading performance. Journal of Occupational and Organizational Psychology, 76: 53-68. Filer, Randall K., Jan Hanousek, & Nauro F. Campos. 1999. Do Stock Markets Promote Economic Growth? Fleming, Michael & Jose A. Lopez. 1999. Heat Waves, Meteor Shower, and Trading Volume: An Analysis of Volatility Spillovers in the U.S. Treasury Market.In Francisco, Federal Reserve Bank of San, editor, Working Papers in Applied Economic Theory San Francisco: Federal Reserve Bank of San Francisco. 37 Fornell, Claes & F.L. Bookstein. 1982. Two Structural Equation Models: LISREL and PLS Applied to Consumer Exit-Voice Theory. Journal of Marketing Research, XIX: 440-52. French, Kenneth R. & James M. Poterba. 1991. Investor Diversification and International Equity Markets. Paper presented at Papers and Proceedings of the Hundred and Third Annual Meeting of the American Economic Association. French, Kenneth R. & James M. Poterba. 1991. Japanese and U.S. Cross-Border Common Stock Investments. NBER Working Paper No. R1537. Gerlach, Stefan, Srichander Ramaswamy, & Michela Scatigna. 2006. 150 Years of Financial Market Volatility. BIS Quarterly Review(September 2006). Ghosh, A., R. Saidi, & K.H. Johnson. 1999. Who Moves the Asia-Pacific Stock Markets-U.S. or Japan? Empirical Evidence Based on the Theory of Cointegration. Financial Review, 34: 159-69. Graham, John L. 1984. The Influence of Culture on the Process of Business Negotiations: An Exploratory Study. Journal of International Business Studies, 16: 81. Hirshleifer, David. 2001. Investor Psychology and Asset Pricing Journal of Finance, 56(4): 1533-98. Hofstede, G. 1984. Culture's Consequences: International Differences in Work-Related Values. Beverly Hills, CA: Sage. Hofstede, G. 2001. Culture's Consequences: International Differences in Work-Related Values. Beverly Hills, CA: Sage. Hofstede, G. 1991. Cultures and Organizations: Software of the Mind. London: McGraw Hill. Hofstede, Geert. 2005. Geert Hofstede Cultural Dimensions. Holton, Glyn A. 2006. Volatility. Hsee, Christopher. 1998. Researching Risk Preference: When It Comes To Cross-Cultural Predictions, Your Best Guess May Not Be Good Enough. Capital Ideas, 1(3). Hsee, Christopher & Elke U. Weber. 1999. Cross-National Differences in Risk Preference and Lay Predictions. Journal of Behavioral Decision Making, 12(2): 165-79. Hulland, John. 1999. Use of partial least squares (PLS) in strategic management research: a review of four recent studies. Strategic Management Journal, 20(2): 195-204. Jansen, W. Jos & Niek J. Nahuis. 2003. The Stock Market and Consumer Confidence: European Evidence. Economic Letters, 79: 89-98. Jegadeesh, Narasimhan & Sheridan Titman. 2001. Probability of Momentum Strategies: An Evaluation of Alternative Explanations. Journal of Finance, 56: 699-720. Jegadeesh, Narasimhan & Sheridan Titman. 1993. Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency. Journal of Finance, 48(1): 65-91. Jorion, Philippe & William N. Goetzmann. 1999. Global Stock Markets in the Twentieth Century. The Journal of Finance, 54(3): 953-80. Kahya, E. 1997. Correlation of Returns in Non-Contemporaneous Markets. Multinational Finance Journal, 1: 123-35. Kwok, Chuck C. Y. 2006. National Culture and Financial Systems. Journal of International Business Studies, 37: 227-47. Langer, E.J. 1975. The illusion of control. Journal of Personality & Social Psychology, 32(2): 311-28. McSweeney, B. 2002. Hofstede’s Model of National Cultural Differences and Consequences: A Triumph of Faith - A Failure of Analysis. Human Relations, 55(1): 89-118. Morgan Stanley Capital International, Inc. 2005. MSCI Equity Indices. 38 Morrin, Maureen, Jacob Jacoby, Gita Venkataramani Johar, Xin He, Alfred Kuss, & David Mazursky. 2002. Taking Stock of Stockbrokers: Exploring Momentum versus Contrarian Investor Strategies and Profiles. Journal of Consumer Research, 29. Morrin, Maureen, Jacob Jacoby, Gita Venkataramani Johar, Xin He, Alfred Kuss, & David Mazursky. 2002. Taking Stock of Stockbrokers: Exploring Momentum versus Contrarian Investor Strategies and Profiles., Journal of Consumer Research: Journal of Consumer Research, Inc. Mullins, Gary E. 2000. Stock Market Volatility: Measures and Results Stevens Point, WI: Central Wisconsin Economic Research Bureau. Naik, Prasad A. & Chih-Ling Tsai. 2000. Partial least squares estimator for single-index models. Journal of Royal Statistical Society Series B, 62(4): 763-72. Nisbett, Richard E. 2003. The Geography of Thought. New York, NY: Free Press. Olsen, Robert A. 1998. Behavioral Finance and Its Implications for Stock-Price Volatility. Financial Analysts Journal, 54(2): 10-18. Peters, Edgar E. 1991. Chaos and Order in the Capital Markets: A New View of Cycles, Prices and Market Volatility. New York: John Wiley & Sons. Ping, R.A. 2005. What is the average variance extracted for a latent variable interaction (or quadratic)? online paper http://home.att.net/~rpingjr/ave1.doc. Poon, Ser-Huang. 2005. A Practical Guide to Forecasting Financial Market Volatility. New York: John Wiley & Sons, Ltd. Ringle, Christian Marc, Sven Wende, & Alexander Will. 2005. SmartPLS. 2.0 (beta) ed. Hamburg, Germany: University of Hamburg. Rouwenhorst, K. Geert. 1998. International Momentum Strategies. Journal of Finance, 53(1): 267-84. Runtian, Jing & John L. Graham. 2006. Regulation of Entry Revisited: How Culture Plays Its Role. Schwert, G. William. 1989. Business Cycles, Financial Crises and Stock Volatility.In Research, National Bureau of Economic, editor. Cambridge, MA: National Bureau of Economic Research. Schwert, G. William. 1997. Stock Market Volatility: Ten Years After the Crash. NBER Working Paper No. 6381. Shefrin, Hersh & Meir Statman. 1993. Behavioral aspects of the design and marketing of financial products. Financial Management (Financial Management Association ), 22(2): 123. Shefrin, Hersh & Meir Statman. 1993. The Disposition to Sell Winners Too Early and Ride Losers Too Long, Advances in Behavioral Finance. New York: Sage. Shiller, Robert. 1984. Stock Prices and Social Dynamics. Brookings Papers on Economic Activity, 2. Shiller, Robert J. 1988. Causes of Changing Financial Market Volatility, Financial Market Volatility. Kansas City: Federal Reserve Bank of Kansas City. Shiller, Robert J. 1993. Market Volatility Cambridge, MA: MIT Press. Shiller, Robert J. 1993. Stock Prices and Social Dynamics.In Thaler, Richard, editor, Advances in Behavioral Finance. New York: Sage. Singleton, Kenneth J. 1989. Disentangling the Effects of Noise and Aggregate Economic Disturbances on Daily Stock Price Volatility: Graduate School of Business, Stanford University. Smith, Roger B. 1990. Market Volatility and Investor Confidence: Report to the Board of Directors of the New York Stock Exchange, Inc. New York: New York Stock Exchange. 39 Stulz, Rene M. 1999. Globalization, Corporate Finance, and the Cost of Capital. Journal of Applied Corporate Finance, 12(3): 8-25. Stulz, René M. & Rohan Williamson. 2002. Culture, Openness, and Finance. Summers, Lawrence H. 1986. Does the Stock Market Rationally Reflect Fundamental Values? The Journal of Finance XLI(3). Tellis, Gerard J., Stefan Stremersch, & Eden Yin. 2003. The International Takeoff of New Products: The Role of Economics, Culture and Country Innovativeness. Marketing Science. Theodossiou, P., E. Kayha, G. Koutmos, & A. Christofi. 1997. Volatility Reversion and Correlation Structure of Returns in Major International Stock Markets. Financial Review, 32: 205-44. Theodossiou, P. & U. Lee. 1993. Mean and Volatility Spillovers Across Major National Stock Markets: Further Empirical Evidence. Journal of Financial Research, 16: 337-50. Weber, Ewe U. & Christopher Hsee. 1998. Cross-cultural Differences in Risk Perception, but Cross-cultural Similarities in Attitudes... Management Science, 44(9): 1205. West, Joel & John L. Graham. 2004. A Linguistic-Based Measure of Cultural Distance and Its Relationship to Managerial Values. Management International Review, 44(3): 239. World Federation of Stock Exchanges. 2006. World Federation of Stock Exchanges Statistics: World Federation of Stock Exchanges. 40