ریسک مالی کشور و عملکرد بازار سهام : مورد امریکا لاتین

We use the Clark [Cross-border investment risk. Euromoney Books (1991a); Euromoney (1991b); Euromoney (1991c)] methodology to estimate the macroeconomic financial risk premium from 1985 to 1997 for six Latin American countries with the largest stock markets, and test whether and to what extent it affects their stock markets’ performance. We find that the macroeconomic financial risk premium is a significant explanatory variable for five of the countries, that accounts for about 12% of annual variations in the stock market indices. The results indicate that there are no country-specific fixed effects and that sensitivity to changes in the financial risk premium is similar for all five countries.

In this paper, we measure a country’s default risk as a financial risk premium and investigate its effect on the performance of the six largest Latin American stock markets, namely: Argentina, Brazil, Chile, Colombia, Mexico and Venezuela. We focus on Latin America because it has a long history of capitalism and reliance on foreign debt to finance its development. In fact, the debt crisis of 1982 was born in Latin America and a large percentage of outstanding Brady bonds is still associated with Latin American borrowers. Consequently, foreign investors are familiar with the region and the financial risks associated with the countries that comprise it. Besides having the largest stock markets in the region, the six countries in our study are also major international borrowers with the high levels of outstanding foreign debt that are necessary for the existence of a financial risk premium. Country default risk refers to the probability that a country will be unable to generate enough foreign exchange to enable its residents, both public and private, to meet interest and principle payments on their foreign debts. It plays a prominent role in the literature on sovereign debt and country creditworthiness (see, for example, Callier, 1985, Cline, 1984 and Feder & Just, 1977; McFadden, Eckaus, Feder, Hajivassiliou, & O’Connell, 1985) and has profound implications for a country’s economic and social well being. First of all, a de jure or de facto default indicates that a country is unable to generate the foreign exchange it requires to maintain internal and external economic equilibrium. Restoration of equilibrium implies relative price changes, resource reallocation, and income redistribution, with resulting painful effects on levels of output, employment and standards of living.1 Besides the costs associated with economic adjustment, debt default also inflicts two other penalties on the country. The first refers to the costs associated with the loss of access to international capital markets (Eaton & Gersovitz, 1981). The second concerns the costs due to direct sanctions such as the elimination of trade credits or the seizure of assets (Bulow & Rogoff, 1989). Domestic stock markets are intimately linked to domestic economic performance.2 They have also become a major source of foreign capital. Thus, when a de jure or de facto default occurs, the domestic stock market will suffer from the effects of economic adjustment. For example, during the Mexican peso crisis, the Mexican stock market index dropped by 38.7% from December 1994 to February 1995. From July 1, 1997 to February 16, 1998 during the Asian crisis, Thai stocks fell by 48.4%, Indonesians by 81.7%, Malaysians by 58.4%, Philippines by 49.2% and Koreans by 63.1%. The Russian crisis of 1998 caused a stock market decline of 41.3% in the month of August alone. Furthermore, foreign capital fleeing the stock market can also exacerbate the balance of payments crisis and increase the pain of economic adjustment. In spite of the interest in emerging stock markets and country risk in general, there are few papers that deal directly with country default risk and stock market performance. The early literature on default was wrapped up in general country risk with respect to sovereign debt and the determinants of default.3 For example, Feder and Just (1977), McFadden et al. (1985), and Callier (1985) focus on the financial variables, Berg and Sachs (1988) on structural variables, Huizinga (1989) on prices and Hajivassiliou (1989) on debt overhang. Others, such as Ozler and Huizinga (1991) concentrate on the exposure of individual banks, Boehmer and Megginson (1990) and Stone (1991) look at a mixture of measurable macroeconomic data and unmeasurable qualitative information, Edwards (1984) and Kutty (1990) use macroeconomic data and Dooley and Stone (1993) take a fiscal approach. Schmidt (1984) compares the ability of different statistical methods to predict default. More recent studies have examined the ability of sovereign credit ratings to explain or forecast emerging market bond returns (e.g., Cantor & Packer, 1996; Erb, Harvey, & Viskanta, 1994;Erb, Harvey, & Viskanta, 1996a). Where the stock market is concerned, there are no papers (to our knowledge) that deal explicitly with macroeconomic default risk. Kim and Mei (2001), Chan and Wei (1996), Cutler, Poterba, and Summers (1989) and Bittlingmayer (1998) consider political risk only and focus on stock market volatility. Erb, Harvey, and Viskanta (1996b) consider the effects of general country risk on emerging and developed stock markets. Other papers, such as Erb, Harvey, and Viskanta (1995), Cosset and Suret (1995), Bekaert (1995) and Bekaert and Harvey (1997) test stock market performance with respect to various indices designed to reflect a general “country risk,” “creditworthiness” or “political risk” index where default risk is but one of many determinants. The upshot of all these studies, as Diamonte, Liew, and Stevens (1996), Martins, Petrov, and Kelly (2001) and Kim and Mei (2001) have pointed out, is that the relationship between country risk and both debt and equity returns is widely accepted. However, despite the apparent importance of country default risk on domestic stock markets and the fact that domestic stock markets have become a major source of foreign capital for developing countries, no work has been done in this field. The first novelty of this study with respect to the outstanding literature, then, is that we focus specifically on country default risk as reflected in the country’s financial risk premium. Besides testing the pertinence of a variable that theory and practice suggest should be important, by focusing on country default risk we avoid many of the shortcomings associated with other, more general measures of country risk. For example, several researchers and practitioners have examined the process of assessing a country’s riskiness (e.g., Haque, Mathieson, & Mark, 1997; Meldrum, 1999, Meldrum, 2000 and Wells, 1997). One point on which they all agree is that this process is based mainly on subjective criteria. Although there are several variables which should be used in this process which are ‘objective’ in the sense that they can be measured objectively (for a list of such variables see: Larr, 1999), there is a large number of factors which require subjective judgement. Such factors include socio-political institutions and because of their subjective nature, some researchers argue that until their link with economic growth is properly investigated, they will remain the weakest link of the rating process (Meldrum, 1999). One attempt to quantify such factors was made by Balkan (1992) who created two variables of political risk (the level of democracy and political instability) and found that they are significant in explaining debt rescheduling, but research in this area is scarce. Where country ratings themselves are concerned, there is little evidence that default risk is a major determinant. Haque, Mathieson, and Mark (1996) find that the ratings provided by the Institutional Investor, Euromoney and the Economist Intelligence Unit are mainly affected by a country’s ratio of nongold foreign exchange reserves to imports, the ratio of the current account balance to GDP, GDP growth and inflation. Cosset and Roy (1991) found that Euromoney’s and the Institutional Investor’s ratings are mainly affected by the country’s GNP per capita, its propensity to invest and its ratio of net foreign debt to exports.4 The authors suggest that developing countries’ policy makers can focus on these variables in order to facilitate their access to the international money markets.5 The second novelty of the paper is that rather than using interest rate spreads or secondary market discounts on bank debt as a proxy for default risk that is found in the sovereign debt literature, we measure default risk for the whole country directly as a financial risk premium, using the Clark, 1991a, Clark, 1991b and Clark, 1991c methodology.6 The advantage of this approach is that it is comprehensive in the sense that it is not tied to an individual borrower or instrument. The problem with rate spreads and secondary market discounts on bank debt as proxies for default risk is that they reflect the risk of the individual instruments and borrowers rather than the overall risk of the country. Disparities in these measures exist even intra country, depending on (i) fees (ii) timing of loan due to the demand and supply conditions of bank credit, (iii) tax treatment of the bonds, and (iv) loan features such as (a) length of the loan, (b) size of the loan, (c) type of borrower—whether private or public, (d) currency of the issue and (e) any options/swaps attached to the loan. Furthermore, the loan market, while a popular source of risk assessment, may fail in many cases to provide adequate risk perception in the case of a short-term crisis. Folkerts-Landau (1985), for example, argued that interest rates charged in the loan market may not reflect the true risk of lending as bank syndicates enjoy significant information gathering, monitoring and enforcement advantages over bond holders. Edwards (1986) concurs with this and suggests that due to the cohesive nature of these syndicates, the existence of cross default clauses typical of loan agreements, the ability to negotiate across borders and guarantees given by their national monetary authorities, default in bank loans is less likely than in bond markets. By focusing on a countrywide risk premium, we avoid these shortcomings and work with a parameter that is more likely to reflect the country’s true overall financial position. The rest of the paper is organized as follows. Section 2 describes the country financial risk premium and how it is calculated. Appendix A and Appendix B present additional details on its construction. Section 3 presents the data and methodology. Section 4 develops the empirical analysis and Section 5 concludes.

In this paper, we estimate the country financial risk premium for six Latin American countries and test its effects on the performance of the stock markets in those countries. The countries we examine are the largest Latin American markets, namely: Argentina, Brazil, Chile, Colombia, Mexico and Venezuela. We find that changes in the financial risk premium explain about 12% of annual variations in the stock market indices for five of the countries in question, Argentina excepted. The coefficients are significant and have the right sign. The results are also robust in that the financial risk premium remains significant when the other important macro explanatory variables suggested in the literature are added. We conclude that the country financial risk premium developed in the paper plays an important role in the performance of the stock markets of the five countries, Brazil, Chile, Colombia, Mexico and Venezuela. These results stand in contrast to those of Erb et al. (1996b), who found no relationship between country risk ratings supplied by professional providers and stock market performance for developing countries. Testing for Argentina suggests that the financial risk premium plays a significant role but this role is influenced by Argentina’s structural organization and the effects of the beef cycle. Interestingly, we find no evidence of country-specific fixed effects across the six countries, which suggests that they all have similar systematic risk or that there is a strong regional or “address” effect at work. Except for Argentina, there is no significant country-specific sensitivity effect either. This suggests that on average the five countries are similar with respect to their relative position vis a vis the rest of the world. A detailed analysis of why this should be so is outside the scope of this paper and will be the focus of our ongoing research. However, we can say that besides being from the same region with a common Spanish/Portuguese heritage, they are all, to a certain degree, at similar stages of economic, financial, social and political development, which might explain the similarities.