پویایی گسترش اعتبار مسائل یوروی آمریکای لاتین در بازار اوراق قرضه بین المللی

This paper investigates two important relationships using the sovereign issues made by major Latin American economies in the international bond market: the determinants of credit spread changes using variables derived from structural and macroeconomic theory and the impact of a default episode on the underlying equilibrium dynamics. We find four significant determinants of credit spread changes: an asset and interest rate factor—consistent with structural models of credit spread pricing; the exchange rate—consistent with macroeconomic determinants and the slope of the yield curve—consistent with a business cycle effect. Also, an intra-regional analysis of sovereign yields reveals a shift in the long-run equilibrium dynamics around the Argentine default on the 23 December 2001.

In the lead up to the financial crisis in 2007–2008 associated with subprime debt there was a persistent tightening of emerging market credit spreads, which at the time raised considerable concerns among different parties including multilateral financial institutions, investors, practitioners, central bankers, as well as academics. In the past, such credit spread tightening has often signaled the onset of economic and financial crises1 induced by the formation of pricing bubbles, or the decoupling of economic and financial fundamentals from the asset pricing process. Emerging economies in general and the highly indebted Latin American economies in particular have indeed witnessed a number of credit events which inevitably led to such crises of these economies in the recent past.2 While different arguments for this persistent tightening of credit spreads have been provided, we do not yet know the precise economic reasons for this behaviour.3 It is the objective of this paper to provide further insights into these important issues by modelling the credit spreads and yields of issues by key Latin American sovereign issuers (Brazil, Chile, Colombia, Mexico and Venezuela). US$ denominated bonds issued by these sovereigns in Eurobond markets represent nearly one-half of the entire US$ outstandings. The sample period covers daily yields from 25 February 2000 to 13 January 2006 and comprises 1483 observations for 18 bonds. Sovereign issues also benefit from being the most liquid and actively traded of this class of securities. Investigating the determinants of emerging market credit spreads and understanding the complex equilibrium dynamics that exist in these markets also allow an understanding of the more general mechanisms underpinning the behaviour of credit spreads. Despite several attempts by researchers to identify the determinants of change in credit spreads, much uncertainty remains (Turnbull, 2005). In addition, much of the existing empirical literature investigating the determinants of credit spreads is limited to bond issues by corporations based in mature financial markets. Although credit spreads incorporate the likelihood of default in relation to a riskless bond, the causes and consequences of default in sovereign settings differs from that of corporate settings.4 Structural models of pricing credit-sensitive instruments in a corporate setting predict asset and interest rate factors are the key determinants of credit spreads (measured by the interest rate difference over a risk-free benchmark, typically a US$ government bond). While there is a general consensus in mature financial markets that these two factors are the main drivers of corporate credit spreads (Collin-Dufresene et al., 2001), the generality of this proposition to emerging markets is limited given the institutional differences. Of particular importance for this study is the critical issue that the applications of structural models, which capture the dynamics of the firm in a risk-neutral setting, have not been adequately applied to sovereign settings. Duffie et al. (2003) argue that the decision by a sovereign issuer to default on its debts is driven largely by political factors, which inevitably have economic and political ramifications, including a loss of reputation. Essentially, institutional arrangements governing bankruptcy laws provide the necessary framework to deal with issues arising from the default of a private economic agent. This allows corporate bond holders recourse to a national bankruptcy code in the event of a default by the issuer. However, there are no such clear-cut procedures governing defaults in sovereign settings. Thus, understanding the forces that drive the credit spreads of these instruments and the dynamic relationships that exist between them remains a critical task in the pricing process. When it comes to pricing risky bonds, it is industry practice to attach a yield spread (credit spread) over a risk-free benchmark to reflect the markets’ assessment of an issuer's credit worthiness and the associated default risk of a given fixed income instrument (De Almeida et al., 1998). Structural models of default,5 such as the model proposed by Longstaff and Schwartz (1995), provide a simple and intuitive framework to capture the factors that drive credit spreads with two main factors identified as driving credit spreads—an asset factor (represented by a stock market index) and an interest rate factor (represented by a mid-maturity risk-free bond). In addition to investigating the determinants of credit spread changes in emerging markets, this investigation also focuses on the change in behaviour of these factors and the equilibrium dynamics around the December 2001 Argentine default. We observe a structural break in our data set following this default and so we investigate the equilibrium dynamics before and after the event. One of the important aspects of understanding the equilibrium dynamics is that the practice of attaching a spread to risk-free bonds assumes an equilibrium relationship between different credit classes of bonds and a linear relationship between risk-free and risky interest rate for the specified maturity (Batten et al., 2000). The Argentine default in December 2001 is an ideal setting to test these relationships since there were no other major defaults in the Latin American region, or in other emerging markets, following the start of our sample period in 2000. In particular we employ a modified version of the Longstaff and Schwartz (1995) structural model that allows for daily data analysis in contrast with the monthly data employed in many earlier applications of these models. The same set of observed data is utilised for the investigation of the long-run dynamics and for a structural break. In summary, in order to understand the behaviour of emerging market credit spreads and to empirically establish the economic reasons for the persistent decline in credit spreads, we investigate three important questions: (i) What are the determinants of credit spread changes of US$ denominated sovereign Eurobonds of Latin America and are these determinants consistent with existing empirical evidence on corporate credit spread behaviour? (ii) What is the nature of the long-term equilibrium relationships that exist in these markets? (iii) What happens to the long-term equilibrium relationship when a default episode is triggered in the region? The paper is organised as follows: In Section 2 we outline the data used in this study and briefly discuss the features of Latin American fixed income markets. Section 3 outlines the method used in this study. We discuss the statistical properties of Latin American spreads in Section 4. Section 5 concludes the paper.

Traditional structural theories of corporate credit spread behaviour predict that asset and interest rate factors are negatively correlated to the changes in credit spreads on risky corporate bonds. This paper investigates this theoretical proposition using the yield spreads on sovereign Eurobonds issued in international markets by major Latin American issuers (Brazil, Chile, Colombia, Mexico and Venezuela). The sample period is from 25 February 2000 to 13 January 2006, which includes the default by Argentina on the 23rd December 2001. We find that the change in credit spread is negatively related to the first two factors while the exchange rate is positively related (consistent with local currency depreciation as credit spreads rise). The slope of the yield curve, which captures the business cycle effect, has a positive relationship with short bonds and a negative relationship with the long bonds in our sample. The findings are broadly consistent with structural theory and other empirical findings for corporate and sovereign bonds from other emerging and developed markets. Overall, these findings suggest that sovereign bonds in international markets and corporate bonds issued in US domestic markets appear to be sensitive to the same interest rate effects. The statistical significance of the exchange rate factor representing country risk and the slope variable representing the business cycle effect may be attributable to a sovereign risk premium demanded by investors to invest in these bonds. We also find significant ARMA terms, which indicate inertia in the price adjustment process of sovereign credit spreads. The results also suggest that there was a change in the equilibrium relationship between these sovereign issuers with a decline in the importance of the cross-border cointegration relationship: intramarket relationships have become more important than intermarket (bivariate or pairwise) relationships. Future research could extend this study to a multivariate cointegration framework enabling the identification of common features that exist across different maturities in emerging markets.