آزادسازی مالی و تغییرات در رفتار دینامیکی نوسانات بازارهای در حال ظهور : شواهدی از چهار بازار سهام در آمریکای لاتین

This paper examines whether the dynamic behaviour of stock market volatility for four Latin American stock markets (Argentina, Brazil, Chile and Mexico) and a mature stock market, that of the US, has changed during the last two decades. This period corresponds to years of significant financial and economic development in these emerging economies during which several financial crises have taken place. We use weekly data for the period January 1988 to July 2006 and we conduct our analysis in two parts. First, using the estimation of a Dynamic Conditional Correlation model we find that the short-term interdependencies between the Latin America stock markets and the developed stock market strengthened during the Asian, Latin American and Russian financial crises of 1997–1998. However, after the initial period of disturbance they eventually returned to almost their initial (relatively low) levels. Second, the estimation of a SWARCH-L model reveals the existence of more than one volatility regime and we detect a significant increased volatility during the period of crisis for all the markets under examination, although the capital flows liberalization process has only caused moderate shifts in volatility.

During the last 15 years we have witnessed substantial development in the structure of both mature and emerging financial markets. The emerging markets have been the focal point of interest of private and institutional investors as well portfolio managers. More specifically this growing interest has arisen from the recognition of the important positive link between the process of financial liberalization which these economies have undergone and economic development.1 The flow of portfolio investments to emerging financial markets has increased from a mere $6.2 billion in 1987 to $37.2 billion in 1992 to a total of $211.6 billion for the period 2000–2006 (BIS, 2006). The flow of these funds has been mainly directed to bonds, certificates of deposit and commercial papers, although during the recent period there has been a major shift towards investing in stocks. Financial liberalization, and especially abolishing capital controls directed to an emerging country's stock market, is the result of the decision by the government of this country to allow foreigners to invest in equity in the country's stock market. Over the recent period several researchers have studied the possible benefits and costs of such liberalization processes. A major argument in favour of such a move is that the opening of financial markets in the emerging economies by removing existing capital controls will help them to attract foreign capital to finance economic growth. Furthermore, these increased capital flows will speed up the development of stock markets which may lead to long-run economic growth. The argument for such a positive relationship is based on the prediction that the liberalization of stock markets reduces the aggregate cost of equity capital. An additional implication is that following the opening of financial markets we should observe a rise in physical investment as a result of the decline in the cost of equity capital. Finally, there will be an urgent need for increased transparency and accountability by the firms’ management and this will also result in improved allocation of resources, reduction in the risk of holding stocks as well as to a reduction in the cost of capital (Henry, 2000 and Kim and Singal, 2000). The most important concern by the governments and policy practitioners of emerging economies is that such abolition of capital markets and the subsequent rise in capital flows may give rise to some unpleasant effects. These uncertainties are the result of the fact that international capital flows are very sensitive to changes in interest rates as well as expectations about future economic growth and expected returns from holding stocks. Such changes, even of a small magnitude, may result in negative effects on the domestic economy. Moreover, financial liberalization will result in exposure to foreign influence, making domestic stock prices more volatile as a result of external impact from global financial markets. A final important concern deals with the prediction that capital inflows will cause an appreciation of the domestic currency, which will cause exports to fall for those countries which are export-oriented, whereas in the case of not enough available investment, plans to absorb the money inflow will result in rising inflation. During the 1980s and early 1990s several Latin American and Asian countries underwent a number of structural reforms, abolition of capital controls and global integration processes. But these processes were accompanied by financial crises such as the 1994 Mexican currency crisis and the more recent 1997–1998 turmoil in East Asian financial markets. A substantial number of papers which have scrutinized the operation of the emerging markets of this period have shown that the most common feature of these markets is the high volatility observed in the returns of financial securities. Given this stylized fact of the emerging markets’ instability, many authors have attempted to evaluate the effect of financial reform on several characteristics of emerging markets (Bekaert and Harvey, 2000, Henry, 2000, Bekaert et al., 2002a and Bekaert et al., 2002b). Furthermore, Bekaert and Harvey, 1995 and Bekaert and Harvey, 1997, De Santis and Imrohoroglu (1997), Huang and Yang (1999), Aggarwal et al. (1999), Kaminsky and Schmukler (2003), Edwards and Susmel (2001), Bekaert et al. (2006), Cunado et al. (2006), Chiang et al. (2007) and Moore and Wang (2007) are examples of studies that have examined the issue of increased volatility in emerging markets following financial liberalization. The evidence from these studies is mixed, implying that there is no definite conclusion that the abolishment of capital controls and the opening of the markets has contributed to an increased degree of uncertainty. This paper focuses on the effects of the financial liberalization which four emerging Latin American economies, namely Argentina, Brazil, Chile and Mexico, implemented in the late 1980s and early 1990s from a rather different perspective by considering the dynamic linkages among stock markets. Specifically, we examine two important issues with respect to short-run dynamics. Firstly, we estimate the conditional relationships between the four Latin American countries and the US market. These conditional relationships are a useful tool to assess whether both types of stock markets are affected by crises such as the 1997 Asian crisis which take place in other regions. Additionally, we adopt Forbes and Rigobon's (2002) argument that, during periods of crisis, an increase has been observed in the corresponding correlations as a result of increased volatility in world markets. This analysis is done by applying the Dynamic Conditional Correlation (DCC) specification developed by Engle (2002) and Engle and Sheppard (2001) to estimate the conditional relationships between the five equity markets. Secondly, we implement the Markov Switching ARCH-L (SWARCH-L) model developed by Hamilton and Susmel (1994) to investigate the existence of structural breaks in volatility of these markets during the period of the financial liberalization process and the emergence of the Mexican peso crisis, during the 1997–1998 Asian and Russian financial crises, and in the aftermath of the 2001 terrorist attack in the US. Understanding these issues leads to several important policy implications, since proponents of capital controls imposition have argued that periods of financial instability are transmitted across countries. The main findings of this paper are summarized as follows. First, the implementation of the DCC shows that the conditional correlation coefficients are relatively low for the majority of the bivariate cases. Second, the conditional correlations rose dramatically during the Asian and Russian financial crises of 1997–1998, and remained at a significantly higher level until the end of the examined period. This finding provides supportive evidence of contagion and is in line with Forbes and Rigobon (2002), who argue that the increased correlations over crises periods are due to an increase in volatility in world markets because of a financial crisis. Fourth, the estimation of a univariate three-state SWARCH model leads to the conclusion that there are episodes of high volatility for all markets, especially around the Asian and Russian financial crises. The rest of the paper is organized as follows. In Section 2, we discuss the financial liberalization process and volatility in Latin American stock markets. Section 3 presents the econometric methodology. In Section 4, we report and discuss empirical results whereas Section 5 gives our summary and concluding remarks. 2. Financial liberalization and stock market volatility Since the mid-1980s, Latin American governments have implemented various programmes of financial liberalization. Table 1A presents the beginning dates of the opening of the markets of the four emerging Latin American markets. These dates are identical or very close across authors, so we can consider them as the appropriate dates for the purpose of our analysis. Table 1B provides alternative signals of liberalization which include the Official Liberalization Date and the dates of introduction of the First Country Fund and First American Depository Receipts (ADR). Table 1C reports various indicators of direct and indirect barriers for institutional investors which are used to assess the extent of liberalization in these economies.2

In the aftermath of the Asian and Russian financial crises of the 1990s, several authors have argued that the high degree of capital mobility has created significant instability in the international financial system. Therefore, it was suggested that the emerging economies should adopt policies to reduce capital mobility. The well-known Tobin tax was thus brought back into the relevant economic policy discussions as a potential mean of reducing world financial instability. In addition propositions in favour of implementing capital controls were once again heard. In this paper we use weekly data of stock returns of four emerging Latin American emerging markets (Argentina, Brazil, Chile and Mexico) as well as of the US market, in order to analyze the existence of short-term relationships between these markets as well as the behaviour of volatility through time. The first part of the analysis involves the estimation of conditional correlations among the five equity markets. To this end we implement the Dynamic Conditional Correlation specification proposed by Engle (2002) and Engle and Sheppard (2001). The second part involves an analysis for the existence of volatility regime switching by applying the Markov Switching ARCH-L model of Hamilton and Susmel (1994) to study for structural breaks in volatility of the examined markets during the examined period. The overall results of the DCC analysis suggest that the conditional correlation coefficients are relatively low for the majority of the bivariate cases. This finding has far reaching implications for portfolio diversification gains in terms of reduction of risk. Moreover, there is substantial evidence that the short-term interdependencies between the Latin American stock markets and the US stock market were affected mostly by crises in other emerging market regions, and particularly the Russian crisis that appeared to have the greatest impact on the conditional correlations. Furthermore, the conditional correlations, apart from the case of Argentina-US, rose dramatically through the period of the recent financial crises and remained at a significantly higher level until the end of the examined period. These results are in line with Forbes and Rigobon (2002) who argue that the increased correlations over crises periods are due to an increase in volatility in world markets because of a financial crisis. Furthermore, we found evidence that there were volatility switching regimes in the Latin American stock markets during the 1990s and early 2000s. The estimation of the appropriate univariate three-state SWARCH models revealed that there were episodes of high volatility for all four markets, especially around the events of financial crises. However, the analysis also made clear that these high volatility regimes were short-lived.