بحران بدهی منطقه یورو : آیا می توان توسعه را از حاشیه به هسته انتقال داد؟

We examine the determinants of joint default risk of euro area countries during 2007–2011. To accomplish this, we recover joint default probabilities from individual CDS contracts. In contrast to earlier theoretical studies, we find that financial linkages are an active contagion transmission channel only in the case of the troubled periphery euro area economies. During the current sovereign debt crisis, real economy linkages play a more important role in transmitting shocks from the euro area periphery towards its core. Countries that have stronger trade interconnections with troubled economies tend to have a higher expected joint default risk

The economic problems of Greece, Ireland, Italy, Portugal and Spain have sparked the fear of contagion that could threaten the sustainability of the euro area (EA). This raises the question of how to measure the degree of vulnerability of sovereigns to default. In this paper, we derive the perceived joint probability of default (JPoD) for EA country pairs, that is, the market consensus regarding the likelihood of extreme negative events to spread among EA countries. Furthermore, we identify the main determinants of increases in the joint probability of sovereign default. We also document what transmission channels might facilitate contagion across the troubled EA periphery and the healthy core during the recent crises. In our setting, contagion is defined as an increase in the joint probability of default after a shock to one country. For most of our analysis, we implicitly assume that the shock originates in the periphery of the EA which consists of Greece, Ireland, Italy, Portugal and Spain. The rest of EA countries are characterized as being part of the EA core. We focus mainly on two potential transmission channels through which shocks from the periphery can transmit to the core: through financial linkages and through real economy interconnections. Financial linkages, defined as the banking claims to and by foreign residents within a country couple, can serve as a transmission channel when banks of one sovereign fall victim to a shock and cannot repay their obligations towards banks of another sovereign. In this case, the latter sovereign might have to step in and recapitalize its banks, incurring immediate losses and damaging its fiscal stance. Real economy interconnections have a similar impact; namely shocks to one country could transmit through the trade credit associated with firms exporting to another country. In this case, firms would incur losses if their counterparts in the other country were to renege on their obligations due to some shock. This would imply lower tax revenues for the sovereigns and decrease their capability of repaying debt. In our measurement of the joint probability of default, we depart from ex-post definitions of default and focus on market expectations regarding such events. In this sense, the sovereign will be considered distressed in our setting whenever the market believes that the sovereign's unobserved assets are below a certain threshold. We recover these expected joint default probabilities from the credit default swap (CDS) spreads of 13 EA countries between January 2007 and August 2011. Our results reveal considerable heterogeneity in the levels of perceived joint default risk across the country pairs that we investigate. However, two events seem to be of major importance for the overall JPoD dynamics: Lehman Brothers' filing for bankruptcy in September 2008 and the Greek government's announcement of its fiscal problems in early November 2009. The subsequent empirical analysis allows us to identify whether shocks were spread through real economy or through financial channels. We find that real economy interconnections (proxied by bilateral trade flows) seem to play a significant role in forming the market perceptions about sovereign default risk, as countries that trade intensively tend to have a higher JPoD during the sovereign debt crisis. Moreover, this channel facilitates the transmission of shocks both within the group of troubled economies mentioned above and from the troubled economies to the rest of the EA countries. We also find that financial linkages are an active contagion transmission channel only in the case of the troubled economies, in contrast to the existing literature suggesting that such interconnections could transmit shocks between EA periphery and core (see e.g. Bolton & Jeanne, 2011). Our results also suggest that stronger and larger economies with a low debt-to-GDP ratio seem to have a lower perceived joint probability of default. Furthermore, we find evidence for changes in the magnitude and significance of these effects across different subgroups of countries, both before and after the aforementioned major events. Regional factors, such as EA market illiquidity and uncertainty, also seem to be important determinants of JPoD across all our specifications. Most recent studies on sovereign distress analyze the linear dependence between countries, instead of focusing on tail risk measures. Longstaff, Pan, Pedersen, and Singleton (2011) conclude that sovereign debt returns are more correlated than equity returns, based on a sample of 26 countries. Pan and Singleton (2008) study the linear dependence of five-year CDS contracts on the sovereign debt of Korea, Turkey and Mexico and find a high level of co-movement between these instruments. Moreover, Reinhart and Rogoff (2011) find that linear dependence during extreme events in sovereign markets seems to be as strong as in normal times. In contrast to these previous studies, we argue that the event of an advanced economy defaulting should be considered a tail event. A well-known fact in the economics and statistics literature is that linear dependence measures such as the correlation coefficient fail to capture the dependence structure in the tail of the joint distribution.1 Therefore, they are inadequate for investigating the current sovereign debt crisis. To overcome this deficiency, we employ a procedure that models the tail behavior of sovereign assets, compatible with the literature on contagion. Our approach is based on the Consistent Information Multivariate Density Optimizing Methodology (CIMDO) developed by Segoviano (2006). This methodology has recently been used by Segoviano and Goodhart (2009) to construct banking stability measures. Under this framework, we view the EA as a joint distribution of its individual constituents. To account for the fat-tail characteristic of financial assets, the CIMDO approach adjusts the probability mass in the tail regions of this distribution with information derived from market data. The benefit of this methodology is that it allows us to model nonlinearities in the multivariate distress–dependence structure, making it more flexible in capturing joint extremes compared to the usual Pearson's correlation coefficient. Furthermore, the CIMDO approach is specifically designed to make efficient use of a limited amount of publicly available time-varying country-specific information. Due to the dynamic updating of the joint density with new empirical information, the underlying dependence structure of the CIMDO distribution is intrinsically time-varying. With the help of this approach, we derive the joint distribution of EA sovereign assets. Focusing on the tails of this distribution yields our sovereign JPoD measure. Sovereign defaults, however rare, entail serious welfare costs not only for the parties involved in the debt contract, but also for third parties if the default risk spreads. On one side, we have the loss of reputation and limited future access to international debt markets of the defaulted sovereign (see Panizza, Sturzenegger, & Zettelmeyer, 2009). On the other side, sovereign defaults have direct negative effects on domestic firms and foreign creditors (see Arteta & Hale, 2008). Thus, sovereign default shocks could easily transmit throughout the EA's financial system due to the complex trade and banking links between EA sovereigns, EA banks and between EA sovereigns and banks. Our country-pair setup allows us to test the importance of financial and trade interconnections in transmitting shocks across countries. After deriving investors' perceptions about joint sovereign default, we undertake an empirical analysis to investigate which of these interconnections are of importance to international investors when analyzing contagion risks. Such an analysis has strong implications for policymakers and regulators alike for a number of reasons. First, policymakers are interested in the level of systemic risk, that is, the risk of a particular negative event to lead to the collapse of the entire financial system. In this respect, our measure of joint default risk provides an estimate of the vulnerability of the EA to a sovereign default event. Other things equal, the higher the (unconditional) joint default probability, the higher the probability of a sovereign to default is, given that another sovereign defaults.2 Second, analyzing market expectations can provide important insights and recommendations for policymakers when deciding which measures should be undertaken to defuse contagion risks. In this context, it is essential to know whether investors perceive the real economy or financial interconnections between EA countries to be of major importance in default spreading across the currency union. Third, as the international financial markets are the main source for public financing in the EA, it is crucial to know what influences investors' decisions when evaluating default risk, since this is a major component in bond pricing, apart from the time value of money and liquidity risk. Recent developments on the CDS and bond markets show that investors do not only price in individual sovereign characteristics and global risk factors (see Hilscher and Nosbusch, 2010, Longstaff et al., 2011 and Pan and Singleton, 2008), but also seem to take into account the interdependence among countries. This paper tries to quantify the influence of the latter. Several theoretical models make clear predictions regarding the relationship between financial linkages and contagion. Bolton and Jeanne (2011) provide a possible explanation on how the banking transmission channel can spur contagion across countries. In their theoretical setup, banks from different countries are allowed to trade bonds of risky and safe sovereigns, thus mimicking a financially integrated union of countries (e.g. the EA). As a consequence, while reducing the cost of default for any financial institution through diversification, this setup generates contagion when the risky country defaults on its debt obligations (transmitting shocks from the periphery to the core in their model). Apart from raising concerns regarding contagion, financial ties play an important role in business cycle synchronization (see Kalemli-Ozcan, Papaioannou, & Perri, 2013), as countries that have stronger financial linkages experience more synchronized cycles during crisis times. Shocks could also spread through real economy channels (i.e. bilateral trade) because of the strong economic links within the common currency area. This channel works via the counterparty risk associated with a trade credit and can cause massive contagion effects both within a country and internationally (see Jorion & Zhang, 2009). Our contribution to the existing literature is three-fold. Our main contribution is that we complement and extend the literature on the identification of determinants of sovereign default risk, which focuses solely on individual default probabilities. To the best of our knowledge, we are the first to analyze the determinants of joint sovereign default risk and the transmission channels of shocks during the sovereign debt crisis. Other studies focus solely on correlations between individual CDS spreads (see e.g. Bai et al., 2012 and Jorion and Zhang, 2007). Second, we are among the first to view a set of countries as a multivariate joint distribution and to examine their joint tail behavior. Some of the few studies with such a setup are Gray, Bodie, and Merton (2007), Segoviano and Goodhart, (2009), and Zhang, Schwaab, and Lucas (2012). None of these previous studies, however, goes beyond descriptive analysis of the probability results. Third, we extend the CIMDO approach at the methodological level, departing from the independence assumption in the previous CIMDO studies that significantly understate the joint distress risk between sovereigns and banks (see Peña and Rodriguez-Moreno, 2013, Segoviano, 2006 and Segoviano and Goodhart, 2009). In addition, we provide a series of robustness checks for the CIMDO methodology that aim at shedding more light on the benefits of this approach. The remainder of the paper is organized as follows. In Section 2, we present our methodology and provide a road map of the paper. Section 3 outlines our JPoD estimation procedure and presents results on the time series of our contagion measure. Section 4 describes the results of our empirical analysis regarding the determinants of perceived sovereign JPoD and the contagion channels during the recent crises. Section 5 concludes.

This paper documents the market-perceived probability of joint default of the EA countries during the subprime and the ongoing sovereign debt crisis. For a long time, the currency union was considered an example of a nearly riskless investment environment, but the recent difficulties of several of its members have threatened its very existence, with fears of contagion being predominant throughout 2010 and 2011. To examine the dependence of sovereign debt issuers in extreme circumstances, we employ a recently developed methodology, the CIMDO approach (Segoviano, 2006 and Segoviano and Goodhart, 2009), to model the default region of the distribution of sovereign assets. Using this approach and the resulting CIMDO distribution, we recover the dynamic path of the sovereign joint probability of default between January 2007 and August 2011. This measure allows us to analyze the probability of detrimental events spreading throughout the EA. Furthermore, we bring some important modifications to the original approach and discuss how these changes affect the JPoD measure itself. We confirm an increase in the joint probabilities of default after two major events — Lehman Brothers filing for bankruptcy in September 2008, and the outbreak of the Greek sovereign debt crisis at the end of 2009. As expected, the economies that are at the center of the sovereign debt crisis, Greece, Ireland, Italy, Portugal and Spain, exhibit high joint probability of default, with noticeable converging paths of this measure after the end of 2009. Next, we identify several major factors that affect the pairwise joint probability of default. This analysis is important from a policy perspective, as a precise identification of the driving forces of joint probability of default could lead to better policies addressing financial stability issues in the EA. Moreover, it extends the existing empirical literature on sovereign debt by identifying the degree of importance of bilateral channels (i.e. trade and financial linkages) in spurring sovereign default risk across countries. Our results suggest that greater economic strength and size decrease perceived joint default risk. Relative indebtedness of sovereigns seems to increase perceived joint default risk, but the effect may not be of economic significance. We find that while the relationship between financial interconnections and JPoD has been negative for our cross-section between 2007 and 2011, this relationship turned out to be insignificant during the sovereign debt crisis. During this period, we find evidence in favor of the transmission of shocks through financial linkages only in the case of the EA periphery sovereigns (GIIPS). Moreover, investors seem to discriminate between this group and core EA countries already during the global recession caused by the collapse of Lehman Brothers in late 2008, as the effects for the latter group are mainly insignificant. We also show that contagion does not spread via the financial interconnections from the GIIPS to the rest of EA countries. We document that countries with stronger trade linkages experienced higher joint probabilities of default. This transmission channel is strongest between the periphery states, but we also find evidence that it may transmit negative spill-over effects to the core countries as well. Future research should further investigate the contradictory impact of the joint resilience potential of sovereigns' joint default risk, since international reserves could serve as a prospective policy tool to be used at the EA level. Another interesting avenue for future work could be quantifying the effect of implicit bailout guarantees on the joint probability of default, as well as the potential effects of the recent ECB interventions on the sovereign debt markets.