بررسی مزایای در معرض خطر صاحبان سهام بازار نوظهور: صنایع، حکومت و عدم قطعیت سیاست های کلان اقتصادی

The average equity risk premium (ERP) in emerging markets is well-known to be significantly higher than in developed markets. But, key reasons for this remain unclear, contributing to investment strategy uncertainty. Here, we use industry-level data for 19 emerging market countries across three regions of the world to first examine the contribution of each industrial stock market to the extra premium paid by emerging markets to international investors from 1995 to present, and then to explore the relative importance of country-level governance and macroeconomic policy uncertainty in explaining both national and regional industry-by-industry ERP behavior. We conduct separate analyses for the emerging market crises period of 1995–2002, and the post-crises period of 2003–2012. Based on both static and dynamic approaches, we find that some industries indeed perform consistently better than others. In particular: (i) the healthcare and basic materials industries mostly contributed to the extra premium paid by the Asian stock market; and (ii) the East European and Latin American stock markets’ extra performances were largely driven by the utilities and consumer services industries, respectively. However, our cross-sectional analyses suggest that country-level governance indicators are not strongly correlated with either national or industry-level returns, with the exception of the consumer goods industry. Lastly, using both rolling-window and DCC-GARCH frameworks, we find that correlations between industrial stock market excess returns and a measure of global economic policy uncertainty are consistently negative, and follow similar patterns. Our empirical evidence as a whole suggests that industrial stock markets are more highly related both within and across countries and regions than has been suggested previously. Contrary to much existing empirical work, our results therefore suggest there is currently little space in emerging markets to exploit cross-industry portfolio diversification benefits.

During the last three decades the emerging market asset pricing and international finance literature has largely focused on the following five issues: (i) the behavior of the emerging markets ERP (i.e. performance measurements); (ii) the predictability of emerging stock markets excess returns (i.e. local vs global information variables, country effects vs industry effects); (iii) the effects of financial liberalization on emerging stock prices, cost of capital and expected returns; (iv) the global integration process in emerging stock markets (i.e. de jure vs de facto integration, country institutional quality, disaster risks); (v) the effects of macroeconomic shocks on emerging economies. 1 Although the average ERP in emerging markets is well-known to be significantly higher than in developed markets, the major reasons for this situation remain widely debated, which contributes to investment strategy uncertainty. A conventional wisdom is that emerging markets compensate investors for the inherent risks in terms of high average returns. It is also largely accepted that the structure of the return distribution of emerging markets is potentially unstable. In other words, ERP tends to be less stable through time in emerging markets than in developed markets. Illiquidity, transaction costs, shaky industrial structures, and political instability have often been seen as potential sources of higher compensation and instability.2 However, existing work has yet to achieve consensus around major drivers of higher ERP in emerging markets. In this paper we examine the behavior of emerging stock market excess returns in an industry-by-industry context, with an aim to clarify the roles of different industrial stock markets in generating higher emerging markets’ ERP. We then undertake a simple assessment of the relationship between governance indicators and average stock market performances on an industry basis, to determine the extent to which country-level governance factors may additionally explain variations in emerging economy industrial stock markets’ average excess returns. Lastly, we use a novel dataset on macroeconomic policy uncertainty to examine the co-movement between the global economic policy uncertainty and the emerging industrial stock market average excess returns.3 The remainder of the introduction section presents a brief review of findings from existing literature for each of these issues, followed by a summary of our major findings. Much of the recent research around emerging market returns focuses on the relative importance of country vs industry effects. A large part of this literature supports the idea that country effects tend to dominate industry effects (i.e. cross-country diversification is more beneficial than cross-industry in a risk-return framework). Serra (2000) finds that emerging markets’ returns are mainly driven by country factors and that cross-market correlation is not affected by the industrial composition of the indices. She argues that geographical diversification dominates, in terms of risk reduction, domestic industrial diversification. Alternatively, Cavaglia et al. (2000) find that industry factors are more important than country factors in the late 1990s, and suggest that diversification across industries might provide greater risk reduction than diversification across countries. For the period 1999–2000, L’Her et al. (2002) obtain a similar result. Brooks and Del Negro (2004), at firm level, using data from January 1990 to February 2002 for 23 developed and 27 emerging markets, observe that the rise in co-movement observed during the late 1990s represent a temporary phenomenon associated with the IT bubble. They explore the evolution over time of country and industry effects outside of the technology, media and telecommunications sectors, and find that there is no significant rise in the importance of global industry effects. They conclude that cross-country diversification is still risk-return beneficial. In contrast, Phylaktis and Xia (2006) show that the industry effects are still dominated by the country effects, suggesting that diversification across countries, especially across emerging economies, is more efficient that diversification across industries. The removal of capital and trade barriers in the last two decades has also motivated several studies which focus on the impact of more liberalized economies on stock prices. Several empirical studies have shown that liberalization has decreased the cost of capital (i.e. expected returns), increased foreign direct investment (FDI), and decreased/increased stock market volatility.4 There also exists a large body of the literature that examines the dynamic linkages between emerging and developed markets in the pre- and post- liberalizations periods.5 While a large part of the literature in the preceding decade treated financial liberalization as a one shot event (i.e. once a stock market becomes liberalized it immediately becomes integrated), most recent studies find that financial liberalization and global integration are not simultaneous processes in emerging markets. For example, Claus and Lucey (2012) examine equity market integration in the Asian Pacific region for the period April–May 2006 and find that financial market liberalization is a necessary but not sufficient condition for stock market integration. In the spirit of Pukthuanthong and Roll, 2009 and Donadelli, 2013b studies the dynamics of the global integration process across emerging markets. Using monthly data from January 1988 to December 2011, he shows that the de jure and the de facto integration are not synchronized. Bekaert et al. (2011) apply a new measure of integration to 69 countries over a sample period of more than 20 years. They find that emerging markets continue to display levels of segmentation above the U.S. benchmark, while developed countries have been effectively integrated since 1993. They also find that financial and trade openness, a country's political risk profile, its stock market development, and the U.S. corporate credit spread (a measure of global risk aversion) are statistically and economically significant in explaining the variation in segmentation. A third line of inquiry has examined the extent to which cross-country ERP differences relate to macro-level indicators of good governance and country stability (such as differences in government and corporate transparency, government effectiveness, rule of law, political stability and corruption), as well as to higher exposure to global macroeconomic risks. Diamond and Verrecchia (1991), among others, have argued that a reduction in informational asymmetry can increase the investment from large investors and hence reduce the cost of capital for the firm. Bushee and Noe (2000) report a positive association between corporate transparency and the volatility of the firm's stock price. Ng and Qian (2004) find that corporate governance is worse in more corrupt countries, lowering firms’ values. Ng (2006) observes that corruption is associated with higher borrowing cost for the firm, lower stock valuation, and worse corporate governance. Gelos and Wei (2006) show that higher corruption decreases investment from foreign investors. Diamonte et al. (1996) show that changes in political risk have larger impact on returns in emerging markets than in developed markets. They also observe that emerging markets have become politically safer over the period 1985–1995. They conclude by arguing that changes in political risk can be used to predict emerging stock returns. In the spirit of Barro, 2006 and Prosperi, 2012 empirically shows that ERP is heavily affected by information frictions deriving from institutional aspects such as corruption, rule of law and quality of government. He finds that the higher degree of corruption in some emerging markets have forced international investors to ask for a higher average ERP. Methodological constraints could also play a role in persisting uncertainty over key drivers of emerging market ERP, in that much of the existing work does not go beyond national level analyses to examine potentially different contributions across industries, nor does it take into account the potential for different drivers of performance across two clearly divergent eras of contrasting emerging market activity before and after 2003 (i.e. emerging market crises). Our research here aims to go beyond these methodological limitations in two ways. First, we employ both static and dynamic elementary modeling approaches to examine the contribution of 10 different industrial stock markets to the higher emerging average ERP observed across emerging economies during the last two decades. To do so, we use an extensive range of data from national and emerging industrial stock markets by focusing on industrial stock market excess returns for 19 emerging countries and three regions. Our approach augments existing work, because while much of the existing literature has been devoted to using industry factors to explain variation across emerging market stock returns, little attention has been paid to understanding variation in industries’ behavior per se. Similarly, existing studies on financial market linkages and financial integration have been mostly based on a single market or a single geographical group of markets (e.g. markets in East Europe, or in Latin America, or in Southern Asia). 6 In addition, the literature around predicting emerging stock returns has mainly focused on national stock market indices. Here, we use a standard performance analysis to explicitly examine the role of industrial stock market indices in generating the observed higher emerging markets’ ERP. Second, we conduct separate analyses for what we term the emerging market crises period of January 1995 to December 2002, and the post-crises period from January 2003 to June 2012, in order to test for potentially different drivers of excess returns across the two eras. This explicit focus on both the emerging market crises period and the post-crises period is relatively unique, because most previous emerging market studies tend to focus either on the post-liberalizations period (early 1990s and onwards), or they uniformly analyze data across the period available. We argue that this is potentially problematic, because the emerging-crises period spans a time of much volatility due to several systemic banking crises in emerging economies, with repercussions for emerging stock market performances and the relative importance of different performance drivers during this period. 7 In contrast, the post-crises period includes an increasing degree of real and financial de facto integration, an increasing co-movement between international stock market returns, and rare but notable economic and political events in developed countries (e.g. subprime mortgage crisis and the European sovereign debt crisis). We argue that efforts to understand more universally applicable drivers of emerging market ERP should treat these two eras separately, and focus primarily on the post-crises period. In terms of our major findings, we find via both static and dynamic approaches that: (i) the healthcare and basic materials industries have mostly contributed to the extra premium paid by the Asian stock market; (ii) the East European and Latin American stock markets’ extra performances have been largely driven by the utilities and consumer services industries, respectively; (iii) average emerging industrial stock market performances are lower than the US industrial stock market performances over the crisis period (i.e. January 1995–December 2002). In a dynamic context, at the industry level, we also show that the importance of the world equity index in explaining variation in national and industrial stock market excess returns is increasing in the post-crises period. That is, the percentage of variance in monthly excess returns explained by the world equity portfolio is increasing over time.8 Our examination of the role of country level governance in explaining varying industrial ERP returns finds only weak evidence for a negative relationship between governance indicators and average stock market performance. However, we do find a statistically significant relationship between the consumer goods industrial stock markets and governance indicators. In contrast, we show that the dynamic unconditional and conditional correlations of global policy uncertainty and emerging industrial stock market returns are consistently negative. We observe that such correlations follow a similar patterns both in the US and in the emerging economies, suggesting particular implications for both country and industry portfolio diversification strategies. In addition, we find that increased stock market volatility increases policy uncertainty and dampens both US and emerging industrial stock market returns. The paper is organized as follows. Section 2 describes the data and presents a preliminary analysis of the data. Section 3 examines the emerging ERP in an industry-by-industry context. Section 4 studies the relationship between governance indicators, and industrials stock market excess returns. Section 5 discusses findings related to the co-movement between US economic policy uncertainty and industrial stock markets. Section 6 concludes.

Over the last twenty years, and especially after liberalization,31 emerging stock markets have captured the attention of many scholars as well as many practitioners. Emerging markets’ empirical regularities are well known (e.g. high returns, high volatility, time-varying moments). Using industry-level data for 19 emerging stock markets across three regions and over two different sub-periods (i.e. crises and post-crises), country governance indicators, and a newly introduced measure of economic policy uncertainty, we improve the existing literature in four main directions. First, we show that some industries contribute more than others in determining the extra premia paid by emerging markets to international investors. In particular: (i) the healthcare and basic materials industries have mostly contributed to the extra premium paid by the Asian stock market; (ii) the East European and Latin American stock markets’ extra performances have been largely driven by the utilities and consumer services industries, respectively; (iii) average emerging industrial stock market performances are lower than the US industrial stock market performances over the emerging market crises period. Second, country-level governance indicators do not explain the cross-section of emerging industrial stock market excess returns during the post-crises period, with the exception of the consumer goods industrial stock market. An alternative explanation, which stems from recent macro-finance literature in an asset pricing consumption-based context, could be that much of the higher emerging average ERP may be explained by higher emerging consumption growth rate volatility (Bansal et al., 2013). Third, we show that the time-varying unconditional and conditional correlations between emerging and US industrial stock market returns and policy uncertainty are consistently negative, suggesting that a higher economic policy uncertainty lowers stock prices. While some emerging markets might still display some degrees of segmentation, their national as well as industrial stock markets seem to follow developed stock markets’ dynamics. In all, the dynamics of the correlation coefficients between emerging and US industrial stock market returns and world equity portfolio return and economic policy uncertainty suggests that cross-industry diversification benefits are negligible. While “crises period findings” point out that portfolios diversification benefits might still be exploited, “post-crises findings” show that industrial stock markets are internationally related, thus, lowering the probability to reduce portfolio risk through cross-industry diversification. In contrast to Brooks and Del Negro (2004), our findings also suggest that the rise in the co-movement across international stock markets of the mid- and late-1990s do not represent a temporary phenomenon (see Figs. C.1–C.3). As documented here and elsewhere, correlations between international stock markets are increasing over time, especially in the last decade (Carrieri et al., 2007 and Donadelli, 2013b). We argue that such results, especially in the post-crises period, are mainly driven by balance sheet linkages among international investors and financial institutions across countries.