WPS7618 Policy Research Working Paper 7618 Impacts of Sovereign Rating on Sub-Sovereign Bond Ratings in Emerging and Developing Economies Sanket Mohapatra Manabu Nose Dilip Ratha Development Economics Global Indicators Group March 2016 Policy Research Working Paper 7618 Abstract This paper explores bond-level, issuer-level, and macro- conditions. Well-developed domestic financial markets also level conditions that affect the distance between sovereign tend to be related to a smaller distance, likely because of credit rating and sub-sovereign debt ratings. Over three- stronger macro-financial links for financial issuers. About quarters of rated foreign-currency sub-sovereign bonds 11 to 26 percent of the bonds had ratings higher than the issued during 1990–2013 in 47 emerging and developing sovereign rating, which was achieved mainly through secu- countries were rated at or below the corresponding sover- ritization structures. This observation is confirmed using eign rating, thus confirming the prevalence of a sovereign a double-hurdle estimation that accounts for bond and ceiling. For bonds rated below the sovereign ceiling, a firm characteristics and macroeconomic conditions. The Tobit regression shows strong sovereign-corporate links sovereign-corporate rating relationship became significantly for financial firms, publicly-owned firms, and local gov- stronger at the peak period of the 2008–09 global financial ernment entities. International bonds tend to be rated crisis, and appears to have weakened in the subsequent years. closer to the sovereign rating during riskier global financial This paper is a product of the Global Indicators Group, Development Economics. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The authors may be contacted at sanketm@iima.ac.in, mnose@imf.org and dratha@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Impacts of Sovereign Rating on Sub-Sovereign Bond Ratings in Emerging and Developing Economies* Sanket Mohapatraa, Manabu Nose† and Dilip Ratha§ a † § Indian Institute of Management, Ahmedabad International Monetary Fund World Bank Keywords: Sovereign credit rating, sub-sovereign rating, international debt markets, fixed income securities, spillover effect, sovereign ceiling violation, global financial crisis JEL Codes: F21, F30, G01, G10, G15. _________________________ * Thanks to Augusto Lopez Claros, Douglas Evanoff, Jamus Jerome Lim, Jayanth Varma, Linda Wu, and seminar participants at the International Atlantic Economic Society conference in Savannah and the 28th Australasian Finance and Banking Conference in Sydney for their useful suggestions on an earlier draft. Thanks to Akash Issar and Sergio Kurlat for excellent research assistance. The authors acknowledge financial support for this work from the World Bank's DEC Research Support Budget and from the Research and Publications Committee of the Indian Institute of Management, Ahmedabad. Contacts: sanketm@iima.ac.in, mnose@imf.org, and dratha@worldbank.org. 1. Introduction The credit rating of an emerging market sovereign is a key factor influencing the ratings of sub- sovereign entities including non-financial corporations, banks, public enterprises, and local governments. Several recent studies have demonstrated the impact of sovereign credit ratings on corporate ratings (Ferri and Liu 2003, Borensztein, Cowan and Valenzuela 2013), on bank credit ratings (Williams et al. 2013, Huang and Shen 2014), and on bank stock returns (Correa et al. 2014). Other related studies have considered the implications of changes in sovereign yield spreads for corporate borrowing costs in developing countries (Durbin and Ng 2005, Dailami 2010, Dittmar and Yuan 2008).1 These suggest a significant spill-over impact from sovereign to corporate ratings, in which the effect is greater for corporates in emerging markets than in developed countries (Ferri and Liu 2003). The sovereign rating puts a “ceiling” on the sub-sovereign ratings in most instances, although the rating of the sub-sovereign entity sometimes exceeded the sovereign rating. Historically, the likelihood of sovereign defaults and the default premia of firms in the same jurisdiction were highly correlated in emerging economies, justifying the practice of “sovereign ceiling” as widely used by rating agencies (Grandes and Peter 2006). Since 2001, however, rating agencies have relaxed this policy with a view that the correlation between the two has been weakened (Moody’s 2005, S&P 2013). Especially, following the 2008-09 global financial crisis, Standard and Poor’s (S&P) reconsidered its sovereign ceiling policy and redefined the maximum notch difference between non-sovereign entity’s rating and the sovereign rating depending on sector and region where the firm operates (S&P 2013). The sovereign rating often serves as a benchmark for the capital-raising activities of both the public and private sectors, particularly when they access international capital markets and issue debt denominated in foreign currencies. In spite of a growing literature on sovereign-corporate linkages, there is very little evidence on how sub-sovereign foreign currency bonds would be                                                              1 In this line of research, Durbin and Ng (2005) found that corporate bond spreads were often lower than the sovereign spread as investors do not always apply the “sovereign ceiling” policy, while Dailami (2010) found that sovereign debt problems tends to add additional capital cost for the corporate bond issuers in emerging market economies. 2   rated relative to the sovereign credit rating.2 In particular, as the sovereign ceiling policy has been relaxed in recent years, it is critical to investigate what determines the distance between the sovereign credit rating and sub-sovereign foreign currency bond ratings.3 Moreover, there has also been little focus on whether the sovereign-corporate relationship materially changed during the 2008-09 global financial crisis and subsequent period characterized by a sharp recovery in emerging economies, highly expansionary monetary policies in advanced economies, and strong debt flows to emerging economies. In this paper, we attempt to fill these gaps in the literature by revisiting the effect of sovereign creditworthiness on sub-sovereign ratings using data on foreign currency bonds issued by sub- sovereign entities in 47 emerging and developing economies during 1990-2013. As discussed in the following sections, we consider the “distance” between the sovereign rating and sub- sovereign debt ratings issued by non-financial and financial issuers, and examine how this distance is related to bond issuers’ ownership structure, sector of issuer, market of issuance, debt characteristics, and macroeconomic and global financial conditions. As an extension of Gourinchas and Obstfeld (2012), we also investigate whether and how the sovereign-corporate relationship changed amid the global financial crisis and why sub-sovereign debts were sometimes rated above the sovereign rating. Finally, we perform robustness tests of our main results by controlling for potential selection bias in debt issuance and including additional controls for firm-specific balance sheet attributes available for a subset of debt issuers. 2. Determinants of sovereign and sub-sovereign linkages Among the various factors which could potentially affect the distance between sovereign credit ratings and sub-sovereign debt ratings, perhaps the most relevant are the differences across financial and non-financial debt issuers, and among publicly-owned and private firms. Due to the strong macro-financial linkages, banks and other financial firms (especially large institutions)                                                              2 The bond rating is an important determinant of the borrowing spread on debt issued by non-financial corporations and banks in international and domestic markets, with better-rated debt benefiting from lower borrowing spreads relative to yields on benchmark securities such as US Treasury bonds. 3 In related work, Williams et al. (2013) and Huang and Shen (2014) found that the impact of sovereign credit ratings on bank credit ratings significantly varies by countries’ macroeconomic factors, while Correa et al. (2014) found that a downgrade of sovereign credit rating has a large negative effect on bank stock returns for those banks that would receive stronger support from their governments. Ağca and Celasun (2012) also outline channels through which higher public external debt (which is typically associated with lower sovereign credit ratings; see Cantor and Packer 1996 and S&P 2013) can raise corporate default premia. 3   typically benefit from explicit and implicit guarantees from the central bank or government in case of financial distress. Hence, debt issued by financial firms is likely to be rated closer to the sovereign rating, compared with debt of non-financial firms. Similarly, the sovereign-corporate relationship would also differ by the ownership structure of firms. Publicly-owned firms in emerging and developing countries, even if nominally free of government control, tend to have privileged access to government support during times of distress. As Gozzi et al. (2012) note, firms issue different types of bonds to raise capital in domestic and international capital markets. Debt characteristics such as whether a bond is collateralized, callable, fixed or floating rate, and currency denomination can influence the sovereign-corporate relationship. For instance, bonds backed by collateral tend to have lower credit risk, which can potentially result in a weaker relationship with sovereign risk. One of the other issues which have received attention is the distinction in risks when accessing foreign debt markets between firms in tradable and non-tradable sectors. Non-tradable firms such as real estate and domestic utilities typically have less access to foreign currency resources, and therefore are more vulnerable during currency depreciations and subject to convertibility risk than tradable firms, making them more dependent on the sovereign. The spill-over impact of sovereign risk on corporate risk could also differ across the markets and currencies in which debt is issued. Although the majority of international sub-sovereign bonds are issued in US dollars, firms in developing countries also issue foreign currency debt in euro, yen and several other foreign currencies (see Section 3). As capital markets of emerging economies mature, local currency debt issuance has also started to increase. Since we intend to use a consistent metric for this cross-country study, we restrict ourselves to only the ratings assigned to foreign currency debt by the three major international agencies, Standard & Poor’s, Moody’s and Fitch. We do not include local currency debt in our analysis for three reasons. First, local currency sub-sovereign debt is typically not subject to sovereign risk, in particular to convertibility and transfer risks.4 Hence, the sovereign ceiling does not appear to apply to local                                                              4Transfer risk refers to the likelihood that a government with foreign debt servicing difficulties imposes foreign exchange payment restrictions on otherwise solvent companies in its jurisdiction, forcing them to default on their own foreign currency obligations. This usually applies to debt issued in foreign currency. Indirect sovereign risk faced by all firms irrespective of currency of debt refers to the likelihood that a firm defaults on its debt as a result of corporates’ financial distress triggered by the country’s economic crisis after a sovereign default (which Standard & Poor’s has called “country risk”). 4   currency bonds. 5 Second, local currency debt is not subject to the problem of ‘original sin’ typically faced by developing countries when they are unable to borrow abroad in their own currencies (Eichengreen, Hausmann, and Panizza 2005). 6 Third, local currency debt is often rated by national rating agencies on rating scales that can differ substantially from those used by international rating agencies for foreign currency debt. Finally, while debt-issue level characteristics remain key determinants of the sovereign-corporate relationship, global economic conditions specifically during the 2008-09 global financial crisis and in the subsequent period would have materially changed the distance between sovereign and sub-sovereign credit ratings. 7 The crisis and post-crisis periods were characterized by large injections of liquidity by central banks in the major advanced and emerging economies as well as more stringent regulatory requirements for the financial sector. The latter may have had an unintended consequence of creating a closer nexus between the sovereign and financial firms. However, looser liquidity conditions and a search for yield by international investors may have resulted in greater interest in investing in sub-sovereign debt, which can potentially weaken the relationship between sovereign and sub-sovereign ratings. The analysis of post-crisis developments relates to a recent study by Basu et al. (2013) that examines the evolution of sovereign credit ratings in the post-2008 crisis period, but our paper extends this analysis to debt markets to consider sovereign and sub-sovereign rating relationship during both the crisis and post-crisis periods. 3. Data description The sample includes long-term foreign currency sovereign credit ratings from the three major international rating agencies, S&P, Moody’s and Fitch Ratings, for a group of 47 emerging and developing economies. Ratings by the same three international agencies for sub-sovereign foreign currency debt and other debt characteristics were obtained from the Dealogic. The                                                              5 Grandes and Peter (2006) estimate the elasticity of corporate spreads with respect to sovereign spreads in South Africa, finding that the sovereign ceiling does not appear to apply to local-currency denominated corporate bonds. 6 The original sin refers to the limited ability of poor countries to issue international debt in their own currencies. In consequence, these countries often resort to borrowing from international markets in the major foreign or “hard” currencies, such as the US dollar, euro, yen and British pound. 7The global financial crisis of 2008 originated in advanced economies, with the collapse of Lehman Brothers in the United States, and subsequently spread to developing countries (Gourinchas and Obstfeld 2012). 5   conversion from letter to numeric ratings on a 1-21 scale, with higher values indicating better ratings, broadly follows Cantor and Packer (1996), Gande and Parsley (2005), and Hill, Brooks and Faff (2010) and is described in table 1. All entities other than those classified as central government and supranational organizations are considered as sub-sovereign entities in this paper.8 These include the non-financial private sector, financial firms (including banks and non-bank financial companies), as well as publicly-owned enterprises and local authorities. Our sample covers 47 emerging and developing economies based on the World Bank’s country classification, which also includes  countries that are currently high income but were considered as emerging countries during the early part of the sample period. 9 Taiwan, China, is excluded as it is not part of World Bank’s country classification and no data are reported in the World Development Indicators. Finally, we exclude countries that have less than three sub-sovereign debt issues rated by at least one of the three rating agencies for the entire sample period (see table 2a). The sample includes debt issued in foreign currencies by sub-sovereign entities. Debt in default is excluded from the sample. Table 2b provides the industry composition of our sample. The largest numbers of sub-sovereign bonds by non-financial firms are in the oil & gas, telecommunications, utility & energy and real estate & property sectors. Among foreign currency sub-sovereign debt issues with rating information, the US dollar accounts for close to 70 percent of the total bonds issued. The euro and Japanese yen together account for another 15 percent, and another 33 currencies account for the remaining 15 percent of rated foreign currency sub- sovereign debt. The final data set is unique as it focuses primarily on foreign currency bond ratings and uses a more comprehensive definition of sub-sovereign issuers than earlier studies (Mizen and Tsoukas,                                                              8In this paper, the category “sub-sovereign” is used in a significantly broader sense compared to the term “sub-national”, which has often been used to refer only to provincial or local authorities. 9The sample broadly corresponds to the World Bank’s classification of low- and middle-income countries, combined with the set of countries included in the MSCI Emerging Markets (EM) Index. We exclude high-income members of the Gulf-Cooperation Council from the sample. Bahamas and Barbados, considered as offshore financial centers and tax havens, were also excluded. Several high-income countries (Israel, Macao, Singapore, Hong Kong SAR, China, and Cyprus) that are not part of the MSCI EM Index were also excluded. 6   2012). The final sample includes 5,033 foreign currency bonds issued by sub-sovereign entities, with 2,185 bonds issued by non-financial issuers and 2,848 by financial firms (table 2b).10 Summary statistics for ratings of foreign currency bond issues for non-financial and financial issuers are presented in tables 3a and 3b. In general, the average sub-sovereign rating is lower than the average sovereign rating across the three rating agencies, although average ratings for financial debt issuers are higher than the average sovereign rating for S&P and Moody’s (table 3a). There is considerable variation in the distribution of sub-sovereign ratings. About 38-43 percent of sub-sovereign bonds had the same rating as sovereign rating and 36-46 percent had lower rating than the sovereign rating (adding up to 74-89 percent of sub-sovereign bonds rated at or below the corresponding sovereign rating). This confirms that the majority of sub-sovereign bonds are subject to a sovereign ceiling. The share of sub-sovereign bonds rated above the sovereign rating is relatively small at 11-26 percent of all bonds (12.3 percent for S&P, 25.9 percent for Moody’s and 11.0 percent for Fitch), with average distance above the sovereign rating of 2.4 to 4.0 notches on our rating scale. The shares of debt issues that are rated above, at and below the sovereign rating differs across financial and non-financial issuers. For financial issuers, nearly half of sub-sovereign debt issues are rated at the sovereign rating, while 52-60 percent of non-financial debt issues are rated below the corresponding sovereign rating. It is worth noting that these summary statistics do not control for the effect of other debt-specific characteristics or country-level or global variables as is done in this paper. The summary statistics for debt characteristics, and macroeconomic and global variables are presented in table 3c. The average issue size is about 160 million in constant 2010 US dollars. About 5 percent of debt issues are securitized by future-flow receivables (about 3 percent of non- financial debt issues and 6 percent of financial sector debt issues). A further 3 percent of debt in both cases is issued by Special Purpose Vehicles (SPVs). Bonds backed by other types of collateral, other than future-flow and SPV structures, constitute another 3 percent of the total. Fixed rate notes account for the bulk (82 percent) of sub-sovereign bonds. About 26 percent of bonds are callable. Non-tradable sectors (such as utilities, real estate, construction, retail, and healthcare) account for 25 percent of non-financial debt issues. Euro and Japanese Yen issues                                                              10Overall, about 85 percent (1.45 trillion in constant 2010 US dollars) of foreign currency sub-sovereign debt for these 47 countries, excluding debt in default and those issued by SPVs, is rated by at least one of the three major international rating agencies. 7   account for 9 percent and 4 percent respectively, with most of the remainder accounted for by bonds denominated in U.S. dollars. Publicly-owned entities account for 37 percent of the overall sample of sub-sovereign bonds, 25 percent for non-financial issuers and 46 percent for financial issuers. The distribution of macroeconomic and global variables is broadly similar across the two types of issuers. Figures 1a-1c show scatterplots of sovereign and sub-sovereign ratings, by public and private issuers, and financial and non-financial issuers for the three rating agencies. These figures illustrate the constrained nature of the data, where most of sub-sovereign ratings are bounded by the sovereign rating, but with several bonds rated higher than the sovereign. These figures also illustrate the strong correlation between the sovereign and sub-sovereign debt ratings for public sector debt issuers, with ratings of public sector sub-sovereign issuers typically close to the sovereign rating (along 45 degree line) compared to the private sector, and for financial issuers compared to non-financial issuers. These are consistent with the earlier discussed notion that public sector enterprises and sub-national authorities, as well as financial issuers, are likely to be closely related to the sovereign. 4. Empirical strategy 4.1. Benchmark model for sub-sovereign debt rated at or below the sovereign rating In the first specification, we consider the sample of sub-sovereign bonds that are rated at or below the corresponding sovereign rating and consider the determinants of the distance of the sub-sovereign rating below the sovereign rating. The ratings for this set of bonds are bounded above by the sovereign rating, known as the “sovereign ceiling” effect. The specification in equation (1) below is estimated using a Tobit model. ∗ Θ (1) ∗ The dependent variable is the difference between the sovereign and sub-sovereign ratings by the same rating agency. For the constrained sample, the observed values of the dependent variable take on a value of zero for sub-sovereign 8   bonds rated at the corresponding sovereign rating and a positive value for bonds rated strictly below the sovereign rating. The baseline specification (1) uses the sovereign and sub-sovereign ratings of the three rating agencies separately to account for the differences in the weights that rating agencies assign to the various explanatory variables.11 Equation (1) includes a vector of sub-sovereign debt issue level controls (Dijt), country-level macroeconomic factors (Mjt), and global financial conditions (Gt). Average marginal effects (AME) which represent the influence of the explanatory variables in determining the distance between sovereign and sub-sovereign debt ratings were calculated for the Tobit model for the relevant groups. Country fixed effects ( , industry dummies ( , and a time dummies ( ) were included in all specifications. 12 In order to account for possible correlation of sub-sovereign debt ratings within countries, the standard errors of the regression were clustered at the level of countries in all regressions. First, the sub-sovereign debt issue level controls (Dijt) include bond-level features such as the log of debt issue size in constant 2010 U.S. dollars, maturity of the debt, and a range of indicators for features of the debt, including whether the bond issue is a fixed or floating rate, whether the bond issue is callable, whether the bond is collateralized, and indicator for the major currencies of issuance (euro, yen) other than the US dollar. The issue size can be a proxy for the importance of the debt issuer in its market and its access to a broad range of financing sources (the so-called “size effect”). For the same borrower, though, a larger debt may carry higher risk and hence result in a lower rating, which we check by adding firm size in robustness tests described below. Bonds with longer maturity may be assigned a lower rating than shorter-term debt as it is easier to predict risk in the nearer term than in the longer term. Since the US dollar accounts for the bulk of foreign-currency bonds, we include dummies for euros and Japanese Yen. As discussed earlier, we also include indicators for fixed vs. floating rate (fixed rate bonds typically carry a higher yield to compensate the bond-holder for holding                                                              11Sovereign ratings by S&P are considered to be the most independent among the rating agencies (Alsakka and ap Gwilym 2010) and S&P is typically more active in assigning ratings (Gande and Parsley 2005). In practice, the ratings assigned by the three major rating agencies tend to be highly correlated, with deviations usually restricted to within one or two notches on our rating scale (see Hill, Brooks and Faff 2010, Ratha, De and Mohapatra 2011). 12The country fixed effects account for time-invariant country level institutional and other characteristics, and the industry dummies for unobserved industry-level factors. 9   the debt to maturity at a given interest rate) and for whether a bond is callable (which gives the issuer the right to prepay the principal prior to the maturity of the bond). An indicator for whether the debt issuer is a state-owned was also included in this vector. For non-financial debt issuers, additional indicators for whether the issuer operates in tradable or non-tradable sectors are included. For financial firms, an indicator for whether the issuer is a bank or a non-financial institution (such as mortgage bankers, security brokers, and other financial intermediaries) was included. Second, the country-level macroeconomic variables include GDP growth, consumer price index (CPI) inflation, private credit-to-GDP ratio, current account balance-to-GDP ratio (with higher values indicating a larger surplus), international trade-to-GDP ratio, and a measure of capital account openness compiled by Chinn and Ito (2006, 2008). GDP growth helps performance of sub-sovereign entities and their ability to service debt obligations, thereby positively impacting a sub-sovereign’s debt rating independently of the sovereign’s rating. High domestic inflation reduces the real value of future earnings and is usually associated with macroeconomic instability, thereby reducing the willingness of sub-sovereign entities to plan future investments. The current account balance captures external risk for the private sector: a high current account deficit indicates large financing needs and increased vulnerability to capital outflows. The share of international trade in GDP is a measure of the level of internationalization of the private sector through trade linkages.13 The credit to the private sector-to-GDP ratio measures the level of financial development. An improvement in access to domestic private financing may weaken the relationship of the rating of foreign currency debt with the sovereign credit rating. However, financial deepening may also strengthen the linkage between the financial sector, macro-economy, and international financial markets. As the financial sector becomes more mature and develops stronger macro-financial linkage, entailing also greater exposure to external shocks, the relationship of the ratings of financial firms with the sovereign rating may become stronger. Similarly, greater capital account openness may weaken the relationship for non-financial firms as they have greater access to foreign capital to service external debt obligations, while for financial firms, capital account                                                              13 The sovereign rating by itself encompasses a range of macroeconomic and institutional variables that capture the risk of default on sovereign debt (Cantor and Packer 1996, Ratha, De and Mohapatra 2011) and therefore was not included directly.  10   openness may be related to greater exposure and more reliance on the national authorities for liquidity support, thereby implying a stronger relationship. Finally, global factors in the Gt vector include the U.S 3-month Treasury bill rate, the U.S 10- year government bond yield, the Chicago Board Options Exchange’s Volatility Index (VIX index), and an indicator for the global financial crisis (GFC) dummy, all at monthly frequency. The US interest rate is used as the benchmark global interest rate given that a majority of global sub-sovereign bond issuance in U.S dollars and for the outsized importance of the U.S Federal Reserve in influencing global monetary and financial conditions (Rey 2013). The 3-month and 10-year rates are included separately as the short-term rate (which are determined mostly by monetary policy), while the long-term rate can be influenced by market expectations. The global financial crisis dummy takes a value of one for a one-year window starting from the third quarter of 2008 until the second quarter of 2009 in the benchmark specifications where the crisis variable is used as an additional control. Global factors such as a low interest rate environment can make it easier for firms to access capital, and hence may widen the distance between the sovereign and sub-sovereign ratings. Conversely, higher global risk can tighten the distance as firms may become dependent on the sovereign during times of turmoil in global financial markets. 4.2 Sub-sovereign debt ratings above the “sovereign ceiling” About 11-27 percent of the bonds in our sample are rated above the corresponding sovereign rating. 14 Borensztein et al (2013) discuss the “sovereign ceiling lite hypothesis”, where corporates can be rated above the sovereign rating, and thus a ceiling does not impose an absolute constraint on corporate ratings. In order to account for the corner solution properties of our data, we use a double-hurdle model (Cragg 1971, Wooldridge 2010) for the determinants of the distance of the sub-sovereign debt rating above the sovereign ceiling. This distance is modeled as a function of debt characteristics, firm-level attributes, and country and global factors, conditional on a first stage equation for whether a bond is rated above the sovereign rating or not. An advantage of the double-hurdle model is that it allows the effect of the                                                              14 The share of bonds rated higher than the sovereign rating is 11.4 percent for those rated by Fitch, 13.6 percent for S&P and 26.6 percent for Moody’s. Separate estimations were implemented for the three rating agencies to allow for differences in the effect of potential explanatory variables across the agencies. 11   covariates to vary across the two stages, unlike the more restrictive Tobit model which assumes a single mechanism. The first stage equation investigates factors affecting the likelihood that a sub-sovereign rating is higher than the sovereign rating using the probit model defined in equation (2). Φ Θ (2) The dependent variable in the probit regression model is a binary variable that takes on a value 1 when the sub-sovereign rating is higher than the sovereign rating and 0 otherwise. i refers to the debt issue, j refers to country, k refers to industry and t refers to time period. Using equation (2) as the first stage, the second stage of the double-hurdle model regresses the distance of the numeric equivalent of the sub-sovereign debt ratings above the sovereign rating on controls for bond issuer’s ownership, debt issue characteristics, country-level macroeconomic factors, and global financial conditions in equation (3) below. ∗ Θ (3) ∗ The dependent variable in the second-stage regression is the difference between the sub-sovereign rating and the corresponding sovereign rating by the same agency. 15 It is assumed to follow a truncated normal distribution. Average marginal effects are calculated for the relevant groups for the probit regression for likelihood of the bond rating being above the sovereign rating, and conditional on being above, for the distance above the sovereign rating.16 Country fixed effects ( , industry dummies ( , and time dummies ( ) were included in all specifications. The standard errors of the regression were clustered at the level of countries in all regressions. The set of explanatory variables used earlier in equation (1), which includes a vector of sub- sovereign debt-level controls (Dijt), country-level macroeconomic factors (Mjt), and global                                                              15These are based on numeric scores for sovereign ratings and sub-sovereign ratings which encompass the rating scale (see table 1 for conversion from letter to numeric scale). Debt in default is excluded from the sample. 16 The linear double-hurdle model was estimated using the churdle program in Stata version 14. Average marginal effects, also known as average partial effects, were calculated as the mean of partial effects across the full sample, instead of partial effect only at the sample means (Wooldridge 2010). 12   financial conditions (Gt), are used in both equations (2) and (3). Additionally, we include an indicator for bonds backed by future-flow receivables as such securitization structures may be designed to leapfrog the sovereign rating (Ketkar and Ratha 2008). We also include an indicator for other Special Purpose Vehicles (“Other SPVs”), which mainly includes bonds involving securitization of existing assets, and an indicator for other collateralized bonds. 4.3. Financial crisis and post-crisis regressions We extend our analysis to explore the evolution of the relationship between sub-sovereign and sovereign ratings during the global financial crisis of 2008 and in the post-crisis period. We modify our benchmark model to the Tobit regression as specified in equation (4) below to explicitly examine how the spillover from sovereign ratings to sub-sovereign debt rating changed during this period. (4) The dependent variable is bounded above by the highest rating of 21, and below by the lowest possible rating of 1. is a vector of crisis dummies for the 2008 global financial crisis and post-crisis period, which are equal to 1 when bonds were issued s periods away from the crisis. The financial crisis event is considered to have reached a peak between September and December 2008 which coincides with the collapse of Lehman Brothers in the United States. We use one-year time window to define the onset of the financial crisis (s=0) as we defined in the benchmark regression to investigate the peak effect of the crisis and the evolution of the sovereign-corporate linkages in the immediate post-crisis period (s=1 for 2009-10) and subsequent years (in 2010-12). Controls include all the debt issue-level, macroeconomic and global variables in the benchmark specification. The coefficient of interest estimates how the strength of the sovereign-corporate linkage changed during the peak period of the global financial crisis and in the subsequent period. The relationship between sovereign and sub-sovereign ratings is expected to be strongest during the crisis and progressively weaker as we move further away from the turmoil of financial crisis. We run this regression separately for sub-sovereign ratings of foreign currency debt by three rating agencies. We exclude bonds issued by public entities (SOEs and local governments) as they are rated close to the sovereign rating even during a normal period, as shown in Figure 1a-1c. 13   5. Results This sub-section discusses the results of the benchmark regression as well as the crisis regression as described in equations (1)-(4) in the previous section. 5.1 Results for sub-sovereign debt ratings below the sovereign rating The results for the specifications for sub-sovereign rating below the sovereign rating are presented in table 4 for all issuers. Separate results for non-financial and financial issuers are presented in tables A.1 and A.2 respectively. Additionally, there were a number of local government issues that were rated either at or below the corresponding sovereign rating, hence an indicator for this set of issuers was included in the specifications for the non-financial issuers. The indicator for financial firms is negatively signed and highly statistically significant across the three agencies, with the coefficient in the range of -0.5 to -1.3 notches (table 4). The closer distance between the sovereign and sub-sovereign debt ratings for financial firms compared to other firms confirms the patterns seen in Figure 1a-1c. The indicator for public sector firms is statistically significant at 1 percent or higher across all the specifications in table 4 for all three agencies, with the coefficient ranging of -1.4 to -1.8 notches, indicating a closer distance to the sovereign ratings. Similar results are obtained across the three agencies when considering non-financial and financial issuers separately in appendix tables A.1 and A.2: -0.9 to -1.8 notches for non-financial issuers, and -1.0 to -1.4 notches for financial issuers. Similarly, the local government indicator is also statistically significant across all agencies, with the coefficient ranging from -1.3 to -1.8 notches. The results for publicly-owned entities are consistent with the patterns observed in figure 1. In the appendix table A.2, the coefficient for banks is positive and significant across all agencies, implying that banks are rated closer to the sovereign compared to other financial firms. Among debt characteristics, log issue size is negatively signed for all specifications, although statistically significant only for non-financial issuers, which suggests that larger issues are rated closer to the sovereign. The coefficient of debt maturity is negatively signed in most cases and statistically significant for financial issuers. Notably, the callable indicator is positively signed and statistically significant across all the specifications for both non-financial and financial 14   issuers, indicating that callable bonds are typically at a greater distance below the sovereign rating. This “penalty” in terms of a larger distance may reflect the nature of the bond contract for such bonds where the issuer has a right to prepay the bond prior to maturity. The collateralization indicator is positively signed for non-financial issuers but insignificant, while being negatively signed and statistically significant across all the three agencies for financial issuers (with coefficient ranging from -0.4 to -0.6 notches). The presence of collateral appears to be more relevant for financial firms rated at or below the sovereign ceiling. For non-financial issuers, the non-tradable indicator is positively signed and statistically significant for S&P and Moody’s, indicating a larger distance below the sovereign rating for sectors that typically lack access to foreign exchange (appendix table A.1). Among the macroeconomic variables, the coefficient for capital account openness is positively signed across all three agencies and statistically significant for S&P and Moody’s for non- financial issuers, and for Moody’s for financial issuers. This suggests that the distance below the rating is likely to be lower in countries with lower capital account openness. This may reflect the lack of external financing sources for non-financial sub-sovereign entities in countries with lower capital account openness, making them more reliant on the sovereign (and vice-versa for firms in open economies). The coefficient of domestic credit-to-GDP ratio is negatively signed, suggesting well-developed domestic financial markets tend to be associated with a smaller distance between sovereign and sub-sovereign ratings likely due to stronger macro-financial linkages. The VIX index is negatively signed and statistically significant in most cases for all issuers (table 4) as well as for non-financial and financial issuers (appendix tables A.1 and A.2). This suggests that periods of heightened global financial market risk (when the VIX index is higher) are associated sub-sovereign ratings being closer to the sovereign rating. We do not find robust relationships for GDP growth, inflation, current account balance and trade-to-GDP ratio. Similarly, for global variables, the 3-month and 10-year yields do not exhibit a relationship with the distance of sub-sovereign debt ratings below the sovereign rating. 5.2 Results for sub-sovereign debt ratings above the sovereign rating The results for the double-hurdle model in equations (2) and (3) are presented in table 5. Separate results for non-financial issuers and financial issuers are presented in appendix tables A.3 and A.4. The first (upper) block of the tables provides the average marginal effects of the 15   first-stage probit equation, which estimates the determinants of the likelihood of the debt rating being higher than the “sovereign ceiling”. The second (lower) block presents the average marginal effects of the determinants of the distance of the sub-sovereign rating above the sovereign rating, conditional on the debt rating being higher than the sovereign rating. The results in table 5 and appendix tables A.3 and A.4 suggest that securitization structures, such as future-flow receivables and SPVs, and collateralization, play an important role in achieving a sub-sovereign debt rating higher than the sovereign rating. The future-flow receivable securitization indicator is notably positive and statistically significant determinant of both the likelihood of being rated above the sovereign rating and the distance above the sovereign rating across all specifications. First, the average marginal effect of future-flow receivables on the probability of rated higher than the sovereign rating (upper block of each table) is always the highest for Moody’s, followed by S&P and Fitch. Second, conditional on being rated higher than the sovereign rating, the distance above the sovereign rating for sub-sovereign debt securitized by future-flow receivables is 1.6, 1.1 and 0.7 notches higher than other bonds for Moody’s, S&P and Fitch respectively for all issuers (lower block of table 5). The corresponding distance above the sovereign ceiling for non-financial issuers and financial issuers are similar, with 1.7, 1.1 and 0.8 notches for Moody’s, S&P and Fitch respectively for non-financial issuers (lower block of appendix table A.3), and 1.8, 1.0 and 0.7 notches for Moody’s, S&P and Fitch respectively for financial issuers (lower block of appendix table A.4). The coefficients are statistically significant at the 1 percent level in both stages, indicating a robust relationship. Other special purpose vehicles (“other SPVs”), which are typically backed by existing assets instead of future-flow receivables, also tend to have a higher likelihood of being above the sovereign ceiling and to achieve a higher distance above the ceiling, although the effect is much smaller than that for future-flow securitizations. In the first-stage regression, the coefficients of the SPV indicator are statistically significant at least at 10% in most cases, with the exception of Moody’s for all issuers (upper block of table 5) and Moody’s and Fitch for non-financial issuers in the first-stage regressions (upper block of tables A.3 and A.4). The coefficient of SPV indicator is highly significant in the second-stage equation for distance of all types of issuers across the three rating agencies. 16   A similar pattern is observed for other collateralized bonds, where the coefficients are positive and statistically significant at 10 percent or higher for S&P and Moody’s in the second-stage equation. The effect size is smaller than the effects for the future-flow and SPV indicators. However, for the first-stage equation, the other collateralized variable is significant only for financial issuers and for all issuers in case of S&P and Moody’s and insignificant for non- financial issuers. Overall, these findings suggest a positive role of securitization structures (future flow-receivables and SPVs) and collateralization, especially for financial issuers, in achieving sub-sovereign ratings above that of the sovereign even after controlling for other debt characteristics and country and global variables. Among the three, the magnitude of the effect of future-flow securitization structures is the largest compared to SPVs and other collateralized bonds. Among the debt characteristics, the coefficient of the logarithm of the bond issue size tends to be positive and significant in both the first and second stage regressions. Such bonds tend to have a greater likelihood of being rated higher than the sovereign rating across the three rating agencies, with statistically significant coefficients for financial issues rated by Moody’s (upper block of table A.4) and for non-financial issues rated by Moody’s and Fitch (upper block of table A.3). In the second stage, they tend to be rated higher than the sovereign rating, although the effect is significant for financial issuers. In addition, the results suggest that other debt-specific variables (maturity, fixed rate, callable, currency of issuance) are not major determinants of both the likelihood and the distance to be rated higher than the sovereign rating. They have relatively small marginal impacts or are insignificant in most cases. In table A.3, the non-tradable dummy is negatively signed across all three agencies, which is significant for S&P and Fitch in the first stage equation. This implies that non-tradable issuers, such as domestic utilities, real estate, construction, retail and healthcare, are less likely to cross the sovereign ceiling. The coefficients of public entities and local (sub-national) government indicators are negative and significant in the first stage equation, showing lower likelihood of being above the sovereign ceiling. As suggested by figure 1, publicly entities and local governments tend to be rated close 17   to the sovereign rating even when they leapfrog the sovereign ceiling. On the other hand, financial sector issuers tend to be rated above the sovereign rating, with greater distance above the sovereign ceiling, although statistically significant at 5 percent only for Moody’s. As discussed earlier, this could reflect market expectations of implicit or explicit support from the country’s central bank. For financial issuers, a bank dummy was added to examine the difference between a bond issued by banks and non-banks (in table A.4), which is negative and significant only for S&P and does not appear to matter for other agencies. Among country-specific variables, the capital account openness variable is negatively signed and significant in the first stage only for financial sector issuers, suggesting that a more open capital account is associated with a smaller likelihood of financial firms rated higher than the sovereign (table A.4). Other country-specific variables such as the private credit-to-GDP ratio, GDP growth, and inflation do not exhibit consistent signs and are insignificant in most specifications. The estimates of the global variables suggest that a looser monetary policy in the U.S. and higher global risk are both associated with higher likelihood of sub-sovereigns rated above the sovereign. The 3-month U.S. policy interest rate (the Fed Funds rate) is negatively signed across most specifications in both the first and second stage equations. The VIX index of global risk is positively signed and statistically significant at 1 percent, mainly allowing non-financial issuers to achieve higher ratings than the sovereign. Overall, the double hurdle estimation suggests that certain debt characteristics – in particular, a bond backed by future-flow receivables, SPVs, and collateralized bonds – raise the probability of crossing the “first hurdle” of the sovereign rating ceiling and increase the distance above the ceiling in most cases. Other debt characteristics, country-specific macroeconomic variables, and global variables serve the role of useful controls but are in most cases themselves not robustly related to the sovereign ceiling. 5.3 Effects of the financial crisis The results of equation (4) are reported separately for S&P (column 1), Moody’s (column 2), and Fitch (column 3) in table 6. Column 4-6 includes additional post-crisis dummies to examine the evolution of the sovereign-corporate linkage in the longer post-crisis period. The results presented in columns 1 and 2 show the positive coefficients for the sovereign rating variable as 18   well as for the interaction terms with GFC dummy for S&P and Moody’s, while the result is not statistically significant for Fitch (column 3). This means that the strength of the sovereign- corporate relationship significantly increased during the peak of the financial crisis. Similar results are obtained when additional post-crisis dummies are included. This finding implies a reversal of a decline in the sovereign-corporate linkage at the onset of the financial crisis, and a closer relationship again during the peak of the 2008 global financial crisis. In addition, in the post-crisis period, we found that the correlation between sovereign and corporate ratings weakened in the subsequent years. This was the period when the US Federal Reserve – along with central banks in advanced economies (the Bank of England, the European Central Bank, and the Bank of Japan) undertook quantitative easing to repair the damaged financial market (Fawley and Neely 2013, Lim and Mohapatra, forthcoming). This result suggests that higher liquidity in the aftermath of 2008 financial crisis drove large-scale capital inflows to emerging markets which helped firms mobilize private credit, leading to weaker linkage with the sovereigns. 6. Robustness checks In this section, we perform two types of robustness tests for our benchmark results reported in tables 4 and 5. First, we use balance sheet data to control for firm-specific financial characteristics for a subset of firms where such data is available. Second, we address the issue of possible selection bias through a two-stage Heckman selection model. 6.1 Additional controls for firm-specific characteristics The data on international bond issuance were matched with firm-level balance sheet data from Bloomberg based on International Securities Identification Numbers (ISINs). The balance sheet variables (such as the log of total assets, debt-to-asset ratio, capital expenditure-to-assets ratio, and return on assets) were included along with the set of debt characteristics, macroeconomic and global variables. The inclusion of the balance-sheet variables causes the sample size to drop for the three agencies. The results are reported in table 7 for the Tobit model and in table 8 for the double-hurdle model, corresponding to the benchmark regressions in tables 4 and 5 respectively. 19   In general, the main baseline results remain robust to the inclusion of the balance sheet variables. For bonds rated below the sovereign ceiling (table 7), firm-level variables such as total assets, profitability and capital expenditure are negatively related to the distance below the sovereign rating, while debt-to-assets ratio (leverage) is positively related. This suggests that larger and more profitable firms are rated closer to the sovereign rating. The results for issue size (negative relationship) and callable bonds (positive relationship) are similar to the earlier benchmark Tobit specifications. The coefficient for collateralized bonds is positive for all three agencies and statistically significant only for Fitch. The financial sector indicator is negative for all three agencies, and statistically significant for Moody’s and Fitch (table 8), similar to the benchmark specification. The coefficient for the public sector is negatively signed for all three agencies and statistically significant for Moody’s and Fitch as in the benchmark specification in table 4. These suggest that the results obtained in the benchmark regressions are fairly robust to the inclusion of balance sheet characteristics of firms. For the double-hurdle model for distance above the sovereign rating (table 8), firm-level variables such as total assets, profitability and capital expenditure are positively related to likelihood of crossing the sovereign ceiling and distance above the sovereign rating. The sign of debt-to-assets ratio is inconsistent and statistically insignificant in most cases. The sign and significance of future-flow receivables indicator and other collateralized indicator are positive and statistically significant in most cases, although the "other SPV" variables drops out due to insufficient observations. The coefficient for the financial sector indicator is positive for two out of three agencies in the first-stage regression (similar to the benchmark specification in table 5), but statistically significant only in the second-stage regression for distance above ceiling for Moody's. These suggest that the results obtained in the benchmark regressions are fairly robust to the inclusion of balance sheet characteristics of firms. 6.2 Accounting for possible selection bias In the analysis above, a large sample of sub-sovereign bonds was used to estimate the influence of sovereign ratings. However, as we only observe ratings when sub-sovereign bonds were issued, the sample may differ from the population of all bonds. We address our missing data using a two-step Heckman’s selection model to check the robustness of our baseline results. Our identification strategy relates to Francis, Aykut, and Tereanu (2014) which reshaped loan and 20   bond-level data into country-quarter panel to correct sample selection bias by estimating the probability of corporate bond/loan issuance in the first stage using the average bond-level characteristics (e.g., average day of issuance in country-quarter pair) as an exclusion restriction. For running the selection regression, we similarly need to collapse the bond-level dataset at higher unit, at industry level for instance, to have observations for each month and country where a borrower did not issue any bonds. We decide to categorize bond issuers into eight industry types (based on their ownership type (public vs. private), industry (financial vs. non-financial), and the governing law they used for issuance, and take average sovereign and sub-sovereign ratings for each country-industry-month pairs to build a balanced panel. Should at least one firm in the same industry for each country issue bonds in a particular month using the same governing law, we observe the distance between the sovereign and sub-sovereign ratings for each country- industry-month pairs. In several of the emerging market countries covered in our sample, only a few foreign currency bonds were issued during 1990-2013. For this reason, we restrict our sample to 22 emerging countries where more than 30 foreign currency bonds were issued during the sample period to minimize sampling bias caused by a large number of censored observations. 17 In the selection equation, we use an institutional quality dummy (i.e., governing law is English common law or French civil law) as an exclusion restriction. La Porta et al (1998) predict that common law system in the British tradition offers stronger investor protection than the system in the French civil law tradition, which in turn, promotes the development of market-based financial systems than bank-based financial systems (Levine and Demirguc-Kunt 1999). Based on these works, English common law is expected to increase the probability of issuing a bond, but not directly affect the sub-sovereign rating itself.18                                                              17The sample for the Heckman selection model includes 22 emerging market countries, including Argentina, Brazil, Chile, China, Columbia, Czech Republic, Hungary, India, Indonesia, Kazakhstan, Malaysia, Mexico, Peru, Philippines, Poland, Romania, Russian Federation, South Africa, the Republic of Korea, Thailand, Turkey, and Ukraine. 18 Of the sub-sovereign bonds that have information on governing law (4,105 out of 4,744 bonds), roughly half (2,229) were issued under English Law, followed by 1,351 bonds issued under New York Law. The pairwise correlation of sub-sovereign debt ratings with English Law is 7 percent for non-financial issuers and 9 percent for financial firms, which suggests that choice of governing law is not associated strongly with sub-sovereign debt ratings. 21   In table 9 and 10, the Heckman selection model for distance both above and below the sovereign rating are estimated separately for S&P (in column (1)), Moody’s (in column (2)), and Fitch (in column (3)). The first stage of the selection model is reported in panel A of each table. The coefficient of our instrument (bond issued under English common law) is positive at around 0.6 for the distance above and below the sovereign, with both significant at the 1 percent level. The second stage results in panel B present similar results to the benchmark specification in table 4 and 5 for each rating agency. The coefficient of public entity dummy is always negative and significant as we observed in the benchmark specification, while the coefficient of financial sector dummy is positive for bonds above the sovereign (table 9) and negative for those below the sovereign (table 10). For sub-sovereign bonds rated below the sovereign ceiling, public entities and financial firms are always rated closer to the sovereigns than other types of firms. Well-developed financial market helps firms to issue foreign currency bonds (demonstrated by positive coefficient of private credit-to-GDP ratio in the first stage), while it generally tightens the linkage with the sovereign as a result of stronger macro-financial linkage. As financial sector becomes intertwined into domestic economic structure, it will pose more systemic risk to the economy, which calls for explicit supports from the sovereign. For other macroeconomic and global variables, higher inflation and external deficit, and riskier global financial condition tend to constrain the issuance of international bonds in the first stage. For those rated below the sovereign, the linkage with the sovereign becomes stronger as the VIX index rises in the second stage as we found in the benchmark regression. In all specifications, the selectivity effect ( ) is positively signed (and significant in many cases) suggesting positive selection, i.e. that the observed sample of sub-sovereign bonds is likely to be rated higher than a (counterfactual) random sample. However, the main results remain unchanged, which suggests that selection issues do not materially change our earlier finding. 7. Conclusion This paper contributes to the literature in two main aspects. First, it focuses on international debt issue ratings, rather than corporate ratings or corporate default premia that have received 22   attention in the literature. Given that our sample includes the universe of foreign currency bond issuance by sub-sovereign entities for 47 emerging and developing economies, the findings based on international debt markets provide an additional useful perspective in understanding sovereign-corporate linkages. This is also relevant for the cost of international capital, since the debt rating can affect borrowing spreads (higher bond ratings are typically associated with lower spreads) when emerging market firms issue debt in international capital markets. Second, our work goes beyond the existing literature on the sovereign ceiling effect, and investigates the factors that determine the distance between the two, including the case when sub-sovereigns are rated above the sovereign. This extension allows us to explore bond-level, issuer-level, or macro- level conditions which determine how far the sub-sovereign rating would be generally positioned above or below the sovereign rating as the benchmark. The findings in this paper suggest that key determinants of the distance between sovereign and sub-sovereign debt ratings in emerging markets are clearly different between the case when bonds are bounded by the sovereign ceiling and the case when the sub-sovereign is rated above the sovereign. For the latter case, the analysis found that bonds are usually structured with some forms of securitization to obtain a higher rating than the sovereign, and the bond-level feature in securitization mainly determines how high the corporate could be rated above the sovereign. In many cases, however, the sub-sovereign bond rating is constrained by the sovereign ceiling. In this traditional case, our analysis suggests that the rating agencies are paying attention to corporate ownership structure and sectoral affiliation when evaluating the creditworthiness of sub-sovereign debt. These main results are also robust to possible sample selection bias in bond issuance. For a subset of non-financial and financial firms for which we have matched international debt issuance to the firms’ balance sheet data, the relationship is robust to the inclusion of balance sheet variables, such as firm size, debt-to-assets ratio, and profitability indicators. Overall, the relationship between sovereign and sub-sovereign debt ratings is stronger for financial firms than for non-financial debt issuers. Among non-financial debt issuers, the distance of sub-sovereign debt issued by publicly-owned firms is much closer to the sovereign compared to private sector firms. The closer link suggests that publicly-owned non-financial firms may benefit from the possibility of government support during times of financial stress. 23   Debt characteristics also seem to matter for this relationship. Notably, when sub-sovereign bonds are rated above the sovereign, the future-flow securitization, SPV, and other securitization structures appear to mainly determine the sovereign-corporate relationship. The rating assigned to foreign currency debt issued by firms in “non-tradable” sectors, i.e., those without meaningful access to foreign currency revenues (such as real estate, construction and domestic utilities), has a similar association with the sovereign’s rating as other firms, although foreign debt issued by these non-tradable firms is rated about 2 notches lower on average. This similar ex-ante rating relationship during foreign currency bond issuance stands in contrast with the findings in the literature that firms in non-tradable sectors that have foreign debt on their balance sheets may be adversely affected during sharp currency depreciations. Riskier global financial conditions are also associated with international bonds being rated closer to the sovereign rating for both non-financial and financial firms. Lack of capital account openness is associated with a smaller distance, reflecting reduced access to foreign financing. Well-developed domestic financial markets also tend to be related to a smaller distance, likely due to stronger macro-financial linkages for financial issuers. Finally, the sovereign-corporate relationship is found to be significantly stronger during the peak of the 2008-09 global financial crisis. The correlation has weakened in the post-crisis period characterized by a rise in global liquidity, reflecting quantitative easing by the central banks of advanced economies which has likely helped firms in emerging markets to mobilize private credit with less dependence on the sovereign. Given the increasing access of emerging market sub-sovereign entities to global debt markets, it is important to note that their cost of foreign-currency debt financing is influenced by sovereign credit ratings, especially when firms are owned by states and during sovereign crisis episodes. Finally, the findings of this study call for further careful assessment of sub-sovereign rating decisions on the basis of fundamental corporate financial factors after incorporating the externalities that stem from country and global risk factors. 24   References Alsakka, R., and O. ap Gwilym (2010). “Leads and lags in sovereign credit ratings”, Journal of Banking and Finance, vol. 34(11), pp. 2614-2626. Ağca, Ş., and O. Celasun (2012). “Sovereign debt and corporate borrowing costs in emerging markets”, Journal of International Economics, vol. 88(1), pp. 198-208. Basu, K., S. De, D. Ratha, and H. Timmer (2013). “Sovereign Ratings in the Post-crisis World: An Analysis of Actual, Shadow, and Relative Risk Ratings”, World Bank Policy Research Working Paper No. 6641. Borensztein E., P. Valenzuela, and K. Cowan (2013). “Sovereign Ceilings 'Lite'? the Impact of Sovereign Ratings on Corporate Ratings in Emerging Market Economies”, Journal of Banking and Finance, vol. 37, pp. 4014-24. Cantor, R., and F. Packer (1996). “Determinants and Impact of Sovereign Credit Ratings”, Federal Reserve Bank of New York Economic Policy Review, October edition, pp. 37-53. Chinn, M. D., and H. Ito (2006). “What Matters for Financial Development? Capital Controls, Institutions, and Interactions”, Journal of Development Economics, vol. 81(1), pp. 163-192. Chinn, M. D., and H. Ito (2008). “A New Measure of Financial Openness.”Journal of Comparative Policy Analysis, vol. 10(3), pp. 309 -322. Correa, R., K. H. Lee, H. Sapriza, and G.A. Suarez (2014). “Sovereign credit risk, banks' government support, and bank stock returns around the world.” Journal of Money, Credit and Banking, vol. 46(s1), pp. 93-121. Cragg, J. G. (1971). “Some statistical models for limited dependent variables with application to the demand for durable goods”, Econometrica, vol. 39(5), pp. 829-844. Dailami, M. (2010). “Sovereign debt distress and corporate spillover impacts”, World Bank Policy Research Working Paper No. 5380. Dittmar, R., and K. Yuan (2008). “Do Sovereign Bonds Benefit Corporate Bonds in Emerging Markets?”, Review of Financial Studies, vol. 21(5), pp. 1983-2014. 25   Durbin, E., and D. Ng (2005). “The sovereign ceiling and emerging market corporate bond spreads”, Journal of International Money and Finance, vol. 24(4), pp. 631-649. Eichengreen, B., R. Hausmann, and U. Panizza (2005). “The pain of original sin”, in B Eichengreen and R. Hausmann (eds), Other people’s money: Debt denomination and financial instability in emerging market economies, pp. 13-47, Chicago University Press, Chicago. Fawley, B.W., and C. Neely (2013). “Four Stories of Quantitative Easing”, Federal Reserve Bank of St. Louis Review, vol. 95(1), pp. 51-88. Ferri, G. and L.G. Liu (2003). “How do global credit-rating agencies rate firms from developing countries?”, Asian Economic Papers, vol. 2(3), pp. 30-56 Francis, J.L., D. Aykut, and E. Tereanu (2014). “The Cost of Private Debt Over the Credit Cycle”, Journal of International Money and Finance, vol. 41, pp. 146-181. Gande, A., and D.C. Parsley (2005). “News spillovers in the sovereign debt market”, Journal of Financial Economics, vol. 75(3), pp. 691-734. Gourinchas, P. O., and M. Obstfeld (2012). “Stories of the Twentieth Century for the Twenty- First.” American Economic Journal: Macroeconomics, vol. 4(1), pp. 226-265. Gozzi, J., Levine, R., Martinez Peria, M., and Schmukler, S. (2015), “How Firms Use Corporate Bond Markets under Financial Globalization”, Journal of Banking and Finance, vol. 58, pp. 532- 551. Grandes, M., and M. Peter (2006), “How important is sovereign risk in determining corporate default premia? The case of South Africa”, Working Paper, Bank for International Settlements. Hill, P., R. Brooks and R. Faff (2010). “Variations in sovereign credit quality assessments across rating agencies”, Journal of Banking & Finance, vol. 34(6), pp. 1327-1343. Huang, Y. L., and C.H. Shen (2014). “The Sovereign Effect on Bank Credit Ratings”, Journal of Financial Services Research, vol. 47(3), pp. 341-379. Ketkar, S., and D. Ratha (2008). “Future-Flow Securitization for Development Finance”, in (eds) S.Ketkar and D. Ratha, Innovative Financing for Development, World Bank. La Porta, R., F. Lopez-de-Silanes, A. Shleifer, and R. Vishny (1998). “Law and Finance”, Journal of Political Economy, vol. 106(6), pp. 1113-1155. 26   Levine, R, and A. Demirguc-Kunt (1999). “Stock Market Development and Financial Intermediaries: Stylized Facts”, World Bank Policy Research Working Paper #1462. Lim, J., and S. Mohapatra (forthcoming). “Quantitative Easing and the Post-Crisis Surge in Financial Flows to Developing Countries”, Journal of International Money and Finance. Mizen, P., and S. Tsoukas (2012). “The response of the external finance premium in Asian corporate bond markets to financial characteristics, financial constraints and two financial crises”, Journal of Banking and Finance, vol. 36(11), pp. 3048-3059. Moody’s Investors Service (2005). Revised policy with respect to country ceilings, November. Ratha, D., P. K. De and S. Mohapatra (2011). “Shadow Sovereign Ratings for Unrated Developing Countries”, World Development, vol. 39(3), pp. 295-307. Rey, H. (2013). “Dilemma not Trilemma, The global cycle and monetary policy independence.” paper presented at the Jackson Hole Symposium, August. Standard and Poor’s (2013). Ratings above the sovereign—corporate and government ratings, S&P report, November. Williams, G., R. Alsakka, and O. ap Gwilym (2013). “The impact of sovereign rating actions on bank ratings in emerging markets”, Journal of Banking and Finance, vol. 37, pp. 563-577. Wooldridge, J.M. (2010). Econometric Analysis of Cross-Section and Panel Data. Second edition, MIT Press, Cambridge, MA. 27   Table 1: Conversion of credit ratings from letter to numeric scale Standard & Poor’s and Moody’s Numeric Grade Fitch Rating rating Investment grade Highest credit quality AAA Aaa 21 AA+ Aa1 20 Very high credit quality AA Aa2 19 AA- A3 18 A+ A1 17 High credit quality A A2 16 A- A3 15 BBB+ Baa1 14 Good credit quality BBB Baa2 13 BBB- Baa3 12 Speculative grade BB+ Ba1 11 Speculative BB Ba2 10 BB- Ba3 9 B+ B1 8 Highly speculative B B2 7 B- B3 6 CCC+ Caa1 5 High default risk CCC Caa2 4 CCC- Caa3 3 Very high default risk CC Ca 2 C .. 1 Notes: Higher numeric values indicating better ratings. Credit ratings of sub-sovereign debt issues and of sovereigns that are in default (rating of “D” for S&P and Fitch and “C” for Moody’s) are excluded. Source: Standard & Poor’s, Moody’s, Fitch, Dealogic 28   Table 2a: Emerging and developing countries with rated sub-sovereign foreign currency bonds issued during 1990-2013 Argentina Croatia Jamaica Peru Thailand Azerbaijan Czech Republic Kazakhstan Philippines Trinidad and Tobago Belarus Dominican Republic Latvia Poland Turkey Botswana Egypt, Arab Rep. Lebanon Romania Ukraine Brazil* Estonia Lithuania Russian Federation Uruguay Bulgaria Georgia Malaysia Slovak Republic Venezuela, RB Chile Guatemala Mexico Slovenia Vietnam China Hungary Mongolia South Africa Colombia India Nigeria Korea, Rep. Costa Rica Indonesia Paraguay Sri Lanka Table 2b: Composition of rated sub-sovereign foreign currency bonds issued during 1990- 2013 Industry Obs. Industry Obs. Industry Obs. Aerospace 10 Food & Beverage 114 Oil & Gas 486 Agribusiness 20 Forestry & Paper 57 Professional Services 2 Auto/Truck 23 Healthcare 4 Real Estate/Property 123 Chemicals 59 Holding Companies 21 Retail 34 Computers & Electronics 31 Insurance 7 Telecommunications 312 Construction/Building 132 Leisure & Recreation 4 Textile 3 Consumer Products 19 Machinery 12 Transportation 110 Defense 1 Metal & Steel 111 Utility & Energy 297 Dining & Lodging 11 Mining 97 Local government 85 Total non-financial 2,185 Banks 1,415 Other Financial 1,433 Total financial 2,848 Total rated FC bonds 5,033 Note: Sub-sovereign foreign currency bond rated by Standard & Poor’s, Moody’s or Fitch. 29   Table 3a: Foreign currency sub-sovereign debt ratings and sovereign ratings in emerging and developing economies Percentiles Mean Std. dev. 25th 50th 75th S&P ratings Sovereign rating 13.5 2.9 12.0 14.0 16.0 Sub-sovereign rating 12.5 3.6 9.0 13.0 16.0 Non-financial issuers 11.2 3.3 8.0 11.0 13.0 Financial issuers 13.8 3.4 12.0 15.0 16.0 Moody’s ratings Sovereign rating 13.7 3.1 12.0 14.0 16.0 Sub-sovereign rating 13.2 3.9 10.0 14.0 16.0 Non-financial issuers 11.5 3.6 8.0 12.0 14.0 Financial issuers 14.3 3.5 12.0 15.0 17.0 Fitch ratings Sovereign rating 14.1 2.9 12.0 14.0 17.0 Sub-sovereign rating 12.8 3.7 10.0 13.0 16.0 Non-financial issuers 11.3 3.0 9.0 12.0 13.0 Financial issuers 13.8 3.7 11.0 14.0 17.0 Notes: Both sovereign and sub-sovereign ratings are on a scale of 1 to 21, with higher values indicating better ratings (see table 1) for Standard & Poor’s, Moody’s and Fitch. Sample covers the 1990-2013 period. 30   Table 3b: Foreign currency sub-sovereign ratings above and below sovereign rating Variable All issuers Non-financial Financial issuers issuers S&P ratings Sub-sovereign = Sovereign (% of total) 41.6 28.7 53.4 Sub-sovereign < Sovereign (% of total) 46.2 56.5 36.6 Avg. distance below sovereign rating 3.3 3.8 2.6 Sub-sovereign > Sovereign (% of total) 12.3 14.8 9.9 Avg. distance above sovereign rating 4.0 3.1 5.3 Moody’s ratings Sub-sovereign = Sovereign (% of total) 38.2 28.6 45.0 Sub-sovereign < Sovereign (% of total) 35.8 51.8 24.8 Avg. distance below sovereign rating 3.6 4.3 2.7 Sub-sovereign > Sovereign (% of total) 25.9 19.6 30.3 Avg. distance above sovereign rating 3.0 3.1 2.9 Fitch ratings Sub-sovereign = Sovereign (% of total) 42.8 28.6 51.2 Sub-sovereign < Sovereign (% of total) 46.2 60.2 37.8 Avg. distance below sovereign rating 3.2 3.3 3.1 Sub-sovereign > Sovereign (% of total) 11.0 11.2 10.9 Avg. distance above sovereign rating 2.4 2.2 2.5 Notes: Both sovereign and sub-sovereign ratings are on a scale of 1 to 21, with higher values indicating better ratings (see table 1) for Standard & Poor’s, Moody’s and Fitch. Sample covers the 1990-2013 period. 31   Table 3c: Summary statistics for debt-characteristics, macroeconomic and global variables Percentiles Variable Mean Std. dev. 25th 50th 75th Debt characteristics1/ Log Issue size in const. USD 18.9 1.22 18.1 19.0 19.8 Maturity (years) 6.88 6.48 3.0 5.0 10.0 Future-flow receivable 0.05 0.21 0.0 0.0 0.0 2/ Other SPV 0.03 0.17 0.0 0.0 0.0 Other collateralized 0.03 0.18 0.0 0.0 0.0 Fixed rate note 0.82 0.38 1.0 1.0 1.0 Callable bond 0.26 0.44 0.0 0.0 1.0 3/ Non-tradable 0.25 0.43 0.0 0.0 1.0 Euro issue 0.09 0.29 0.0 0.0 0.0 Yen issue 0.04 0.20 0.0 0.0 0.0 Public entity 0.37 0.48 0.0 0.0 1.0 Macroeconomic variables1/ GDP growth 4.6 3.2 2.8 4.5 6.5 Private credit/GDP 88.6 50.6 41.9 75.3 143.6 Inflation rate 5.97 6.24 2.8 4.4 7.0 Cur. ac. bal./GDP 0.5 4.6 -2.2 0.7 3.5 Trade/GDP 69.5 33.8 50.3 60.9 94.9 Cap. ac. openness 0.5 0.3 0.4 0.5 0.7 Global variables1/ US 3-month int. rate 2.0 2.1 0.1 0.9 4.6 US 10-yr bond yield 3.8 1.4 2.7 3.9 4.7 VIX index 18.6 6.2 14.0 17.3 21.2 Notes: 1/ The sample for debt characteristics, macroeconomic variables and global variables is across the three rating agencies. 2/ Other Special Purpose Vehicles (SPVs) excludes future-flow receivables securitizations. 3/ Only for non-financial firms 32   Table 4: Sub-sovereign debt ratings below the sovereign rating: Results of the Tobit model for all issuers S&P Moody’s Fitch (1) (2) (3) Coeff. Std. Err. Coeff. Std. Err. Coeff. Std. Err. Average marginal effects Distance below ceiling -0.089 (0.107) -0.168 (0.108) -0.210 (0.138) Log Issue size -0.014* (0.008) -0.025** (0.010) -0.017 (0.012) Maturity (years) 0.333 (0.224) 0.605** (0.247) 0.534* (0.282) Fixed rate note 0.915*** (0.128) 0.811*** (0.123) 0.700*** (0.109) Callable bond -0.064 (0.234) -0.020 (0.215) 0.022 (0.249) Collateralized -0.027 (0.164) 0.074 (0.190) 0.512*** (0.192) Euro issue -0.088 (0.112) -0.343*** (0.091) -0.591*** (0.142) Yen issue -1.445*** (0.141) -1.453*** (0.215) -1.843*** (0.194) Public entity -0.880*** (0.155) -1.301*** (0.264) -0.602*** (0.170) Financial sector -0.005 (0.036) -0.015 (0.025) 0.012 (0.032) GDP growth -0.018*** (0.006) -0.016*** (0.005) 0.004 (0.006) Private credit/GDP -0.028 (0.017) -0.025** (0.012) -0.006 (0.035) Inflation rate -0.022 (0.037) 0.021 (0.020) 0.012 (0.023) Cur. ac. bal./GDP -0.010 (0.008) -0.014** (0.007) -0.001 (0.005) Trade/GDP 1.193** (0.556) 1.189*** (0.352) -0.089 (0.754) Cap. ac. openness 0.191* (0.102) 0.117 (0.086) 0.049 (0.137) US 3-month int. rate 0.052 (0.063) 0.152 (0.097) 0.102 (0.086) US 10-yr bond yield -0.029*** (0.007) -0.041*** (0.008) -0.022* (0.013) VIX index 0.377*** (0.062) 0.038 (0.045) 0.265*** (0.098) Constant Y Y Y Country & Ind. dummies 0.198 0.252 0.191 Pseudo R-square -4997.2 -3826.7 -2738.0 Log likelihood 3,431 (44) 2,898 (44) 1,925 (38) No. of obs. (countries) Notes: The dependent variable is the distance of the sub-sovereign debt rating below the sovereign rating. All regressions include country, industry and time dummies. Heteroscedasticity-consistent robust standard errors are clustered at the country level. * Significant at 10%; ** significant at 5%; *** significant at 1% 33   Table 5: Sub-sovereign debt ratings above the sovereign rating: Results of the Double- Hurdle model for all issuers S&P Moody’s Fitch (1) (2) (3) Coeff. Std. Err. Coeff. Std. Err. Coeff. Std. Err. Average marginal effects I(above ceiling) 0.008 (0.008) 0.026* (0.013) 0.005 (0.010) Log Issue size 0.001* (0.001) 0.003*** (0.001) 0.002 (0.001) Maturity (years) 0.263*** (0.025) 0.342*** (0.081) 0.262*** (0.042) Future-flow receivable 0.090** (0.039) 0.064 (0.074) 0.086** (0.039) Other SPV 0.049* (0.028) 0.187*** (0.056) -0.031 (0.052) Other collateralized -0.028** (0.013) -0.030 (0.020) -0.025 (0.026) Fixed rate note -0.005 (0.026) -0.031 (0.030) 0.030* (0.018) Callable bond 0.008 (0.024) -0.013 (0.036) 0.010 (0.027) Euro issue 0.006 (0.036) 0.046 (0.031) -0.071 (0.049) Yen issue -0.023 (0.017) 0.039 (0.048) -0.080** (0.037) Public entity 0.016 (0.029) 0.165*** (0.035) 0.053 (0.044) Financial sector 0.004 (0.003) 0.010 (0.006) -0.006** (0.003) GDP growth 0.001* (0.000) 0.001 (0.001) 0.000 (0.000) Private credit/GDP -0.001 (0.001) 0.000 (0.004) -0.004 (0.004) Inflation rate 0.005** (0.002) 0.008*** (0.003) -0.003 (0.003) Cur. ac. bal./GDP -0.002*** (0.000) -0.003*** (0.001) 0.000 (0.001) Trade/GDP 0.006 (0.039) -0.134* (0.077) 0.089 (0.084) Cap. ac. openness -0.022 (0.014) -0.057*** (0.018) 0.000 (0.011) US 3-month int. rate -0.004 (0.010) -0.012 (0.016) -0.001 (0.018) US 10-yr bond yield 0.002** (0.001) 0.006*** (0.002) 0.001 (0.002) VIX index 0.123*** (0.002) 0.262*** (0.004) 0.112*** (0.011) Constant Distance above ceiling 0.047* (0.025) 0.057 (0.041) 0.008 (0.022) Log Issue size 0.008*** (0.003) -0.001 (0.005) 0.002 (0.002) Maturity (years) 1.089*** (0.078) 1.620*** (0.213) 0.817*** (0.118) Future-flow receivable 0.552*** (0.119) 0.851*** (0.316) 0.457*** (0.129) Other SPV 0.286*** (0.106) 0.693*** (0.167) 0.089 (0.108) Other collateralized -0.195*** (0.058) -0.103 (0.068) -0.032 (0.065) Fixed rate note -0.036 (0.053) -0.148*** (0.046) -0.002 (0.039) Callable bond -0.101 (0.108) -0.014 (0.096) 0.059 (0.057) Euro issue 0.122 (0.185) 0.187* (0.099) -0.073 (0.059) Yen issue -0.376*** (0.102) 0.109 (0.146) -0.131 (0.095) Public entity 0.136 (0.147) 0.436*** (0.158) -0.015 (0.103) Financial sector 0.008 (0.013) -0.006 (0.019) -0.022** (0.008) GDP growth 0.001 (0.001) 0.001 (0.003) 0.000 (0.001) Private credit/GDP -0.004 (0.003) -0.004 (0.008) -0.004 (0.006) Inflation rate 0.011 (0.013) 0.015 (0.011) 0.003 (0.010) Cur. ac. bal./GDP -0.007*** (0.002) -0.011*** (0.004) -0.001 (0.001) Trade/GDP 34   -0.210 (0.137) -0.348 (0.228) 0.186 (0.181) Cap. ac. openness -0.101* (0.053) -0.166*** (0.050) 0.009 (0.025) US 3-month int. rate 0.072 (0.053) 0.176*** (0.068) 0.003 (0.069) US 10-yr bond yield 0.007 (0.005) 0.015*** (0.005) 0.000 (0.007) VIX index 0.495*** (0.029) 0.790*** (0.044) 0.272*** (0.033) Constant Y Y Y Country & Ind. dummies 0.307 0.239 0.292 Pseudo R-square -1812.1 -3280.9 -853.2 Log likelihood 3,914 (45) 3,807 (44) 2,167 (40) No. of obs. (countries) Notes: The dependent variable in the first block is an indicator that the sub-sovereign debt rating is higher than the sovereign rating, and dependent variable in the second block is the distance between sub-sovereign debt rating and the sovereign ratings. All regressions include country, industry and time dummies. Heteroscedasticity-consistent robust standard errors are clustered at the country level. * Significant at 10%; ** significant at 5%; *** significant at 1%  35   Table 6: Sovereign ratings and Sub-sovereign debt ratings during the global financial crisis and post-crisis periods: Tobit Model All issuers excl. public entities S&P Moody's Fitch S&P Moody's Fitch (4) (5) (6) Coeff. Std. Err Coeff. Std. Err Coeff. Std. Err Coeff. Std. Err Coeff. Std. Err Coeff. Std. Err Average marginal effects Sovereign Rating 0.105 (0.076) 0.409*** (0.087) 0.025 (0.107) 0.101 (0.075) 0.424*** (0.084) 0.004 (0.102) Sov. Rating*GFC Dummy 0.157* (0.087) 0.261** (0.132) 0.022 (0.086) 0.169* (0.089) 0.291** (0.135) 0.068 (0.093) Sov. Rating*Post-Crisis 2009/10 0.002 (0.067) -0.243* (0.142) 0.027 (0.050) 0.006 (0.069) -0.280** (0.136) -0.052 (0.056) Sov. Rating*Post-Crisis 2010/11 0.033 (0.048) 0.066 (0.057) 0.143** (0.065) Sov. Rating*Post-Crisis 2011/12 0.018 (0.063) -0.097* (0.052) -0.209*** (0.071) GFC Dummy -2.374* (1.274) -3.386* (1.851) -0.052 (1.242) -2.783** (1.328) -4.337** (1.913) -0.907 (1.357) Post-Crisis 2009/10 -0.054 (1.023) 3.446* (2.070) -0.518 (0.685) -0.115 (1.080) 3.781* (2.000) 0.333 (0.830) Post-Crisis 2010/11 -1.023 (0.695) -1.833** (0.895) -2.204** (0.863) Post-Crisis 2011/12 -0.196 (0.882) 1.042 (0.796) 2.422** (0.956) Other controls 1/ Y Y Y Y Y Y Pseudo-R square 0.185 0.186 0.233 0.185 0.188 0.236 Log-Likelihood -5191.2 -5369.9 -2565.5 -5186.9 -5355.1 -2557.8 No. of obs. 2,337 2,394 1,258 2,337 2,394 1,258 Notes: The dependent variable is the average numeric sub-sovereign debt rating of the three rating agencies. All regressions include country, industry and time dummies. Heteroscedasticity-consistent robust standard errors are clustered at the industry. * Significant at 10%; ** significant at 5%; *** significant at 1% 1/ Control variables in the benchmark specification in table 4 are included, but are not reported. 36   Table 7: Effect of firm-level balance sheet variables on the distance of sub-sovereign debt ratings below the sovereign rating: Tobit model for all issuers S&P Moody’s Fitch (1) (2) (3) Coeff. Std. Err. Coeff. Std. Err. Coeff. Std. Err. Average marginal effects Distance below ceiling Log(Assets) -0.361*** (0.068) -0.350*** (0.081) -0.412*** (0.068) 0.021 (0.014) 0.003 (0.013) 0.015 (0.014) Debt/Assets -1.125 (1.981) -2.956 (2.023) -0.871 (2.474) Capital exp./Assets -0.032* (0.017) -0.026 (0.017) -0.029* (0.015) Return on asset -0.062 (0.094) -0.029 (0.096) -0.214 (0.142) Log Issue size -0.001 (0.009) -0.015 (0.015) 0.002 (0.012) Maturity (years) 0.103 (0.238) 0.461* (0.254) 0.493 (0.329) Fixed rate note 0.495*** (0.149) 0.565*** (0.178) 0.416*** (0.149) Callable 0.357 (0.361) 0.047 (0.507) 1.194*** (0.435) Collateralized 0.197 (0.250) 0.057 (0.328) 0.485* (0.293) Euro issue -0.343** (0.152) -0.415 (0.257) -0.522 (0.339) Yen issue -0.511* (0.286) -0.872** (0.415) -1.146*** (0.222) Public entity -0.685 (0.435) -0.927** (0.419) -0.571 (0.425) Financial Sector 0.010 (0.051) -0.024 (0.036) -0.007 (0.062) GDP growth -0.011 (0.009) -0.004 (0.012) 0.011 (0.009) Private credit/GDP -0.009 (0.023) 0.009 (0.030) -0.038 (0.056) Inflation rate -0.036 (0.053) 0.040 (0.049) 0.061** (0.028) Cur. ac. bal./GDP -0.021* (0.011) -0.025** (0.011) -0.003 (0.007) Trade/GDP 0.856 (0.855) 0.932 (0.780) 0.650 (0.933) Cap. ac. openness 0.082 (0.163) -0.239 (0.206) -0.062 (0.155) US 3-month int. rate 0.112 (0.172) 0.315** (0.133) 0.332** (0.138) US 10-yr bond yield -0.036** (0.017) -0.054*** (0.014) -0.018 (0.015) VIX index 1.486*** (0.049) 0.833*** (0.074) 1.310*** (0.122) Constant Y Y Y Country & Ind. dummies 0.225 0.268 0.187 Pseudo R-squared -2193.0 -1644.5 -1209.0 Log likelihood 1,364 (32) 1,133 (32) 755 (30) No. of obs. (countries) Notes: The dependent variable is the distance between the sub-sovereign debt rating and the sovereign rating. All regressions include country, industry and time dummies. Heteroscedasticity-consistent robust standard errors are clustered at the country. * Significant at 10%; ** significant at 5%; *** significant at 1% Table 8: Effect of firm-level balance sheet variables on the distance of sub-sovereign debt ratings above the sovereign rating: Double Hurdle model for all issuers S&P Moody’s Fitch (1) (2) (3) Coeff. Std. Err. Coeff. Std. Err. Coeff. Std. Err. Average marginal effects Prob.(above ceiling) Log(Assets) 0.028*** (0.006) 0.131*** (0.013) 0.061*** (0.015) -0.002 (0.002) -0.010*** (0.003) 0.003 (0.002) Debt/Assets 0.348* (0.178) 0.318 (0.369) 0.956** (0.458) Capital exp./Assets 0.007*** (0.002) 0.004 (0.004) 0.007** (0.003) Return on asset 0.005 (0.005) 0.010 (0.015) 0.005 (0.014) Log Issue size 0.000 (0.001) -0.001 (0.001) -0.002 (0.001) Maturity (years) 0.347*** (0.067) 0.356** (0.139) 0.474*** (0.088) Future-flow receivable 0.077 (0.053) 0.449*** (0.113) 0.022 (0.081) Other collateralized -0.019 (0.023) 0.014 (0.032) -0.055 (0.039) Fixed rate note 0.006 (0.025) 0.068 (0.045) 0.070** (0.029) Callable 0.030* (0.017) -0.094* (0.050) -0.040 (0.038) Euro issue -0.004 (0.020) 0.025 (0.026) -0.056 (0.036) Yen issue -0.043* (0.026) -0.032 (0.042) -0.141** (0.057) Public entity -0.031 (0.043) 0.112** (0.054) -0.025 (0.059) Financial Sector 0.006 (0.004) 0.019** (0.009) -0.003 (0.007) GDP growth 0.001* (0.000) -0.002** (0.001) -0.001 (0.001) Private credit/GDP -0.005*** (0.001) 0.003 (0.005) 0.001 (0.008) Inflation rate 0.002 (0.006) -0.017*** (0.006) -0.012*** (0.004) Cur. ac. bal./GDP -0.002*** (0.001) -0.001 (0.001) 0.001 (0.001) Trade/GDP 0.036 (0.032) -0.177 (0.160) 0.092 (0.136) Cap. ac. openness -0.028 (0.025) -0.061** (0.025) -0.009 (0.028) US 3-month int. rate -0.017 (0.018) 0.077*** (0.020) -0.033 (0.030) US 10-yr bond yield 0.003** (0.002) 0.006*** (0.002) -0.007** (0.003) VIX index 0.105*** (0.009) 0.302*** (0.026) 0.156*** (0.018) Constant Distance above ceiling Log(Assets) 0.054*** (0.018) 0.209*** (0.021) 0.099*** (0.032) 0.001 (0.008) -0.026*** (0.010) 0.005* (0.003) Debt/Assets 1.107** (0.475) 2.173** (1.108) 2.602*** (0.992) Capital exp./Assets 0.022*** (0.007) 0.004 (0.007) 0.002 (0.010) Return on asset 0.047** (0.019) 0.055 (0.037) 0.010 (0.026) Log Issue size 0.002* (0.001) -0.003 (0.003) -0.003 (0.003) Maturity (years) 1.121*** (0.225) 1.760*** (0.298) 1.280*** (0.238) Future-flow receivable 0.420*** (0.108) 1.113*** (0.202) 0.118 (0.105) Other collateralized 38   -0.088 (0.057) 0.04 (0.063) -0.136** (0.060) Fixed rate note -0.006 (0.043) 0.088 (0.115) 0.100 (0.066) Callable 0.069* (0.039) -0.112 (0.108) -0.051 (0.064) Euro issue 0.113* (0.061) 0.230*** (0.051) -0.033 (0.057) Yen issue -0.025 (0.051) -0.066 (0.109) -0.075 (0.114) Public entity -0.117 (0.113) 0.368*** (0.109) -0.113 (0.121) Financial Sector -0.002 (0.013) -0.014 (0.023) 0.007 (0.010) GDP growth 0.000 (0.001) -0.004* (0.002) -0.001 (0.002) Private credit/GDP -0.011* (0.007) 0.000 (0.015) 0.017 (0.013) Inflation rate 0.001 (0.013) -0.035*** (0.011) -0.010 (0.010) Cur. ac. bal./GDP -0.003*** (0.001) -0.006** (0.003) 0.002 (0.001) Trade/GDP -0.053 (0.098) -0.538* (0.298) 0.247 (0.226) Cap. ac. openness -0.063 (0.052) -0.109 (0.069) -0.017 (0.049) US 3-month int. rate -0.005 (0.040) 0.278*** (0.104) -0.073 (0.052) US 10-yr bond yield 0.004 (0.004) 0.012** (0.006) -0.011 (0.007) VIX index 0.282*** (0.035) 0.622*** (0.058) 0.280*** (0.039) Constant Y Y Y Other controls 1/ Y Y Y Country & Ind. dummies 0.370 0.292 0.394 Pseudo R-squared -483.8 -1197.5 -354.7 Log likelihood 1,429 (32) 1,504 (34) 882 (31) No. of obs. (countries) Notes: 1/ Macroeconomic and global control variables in benchmark specification in table 4 are included, but are not reported. The dependent variable in the first block is a binary variable for the sub-sovereign debt rating being higher than the sovereign rating. The dependent variable in the second block is the distance between the sub-sovereign debt rating and the sovereign rating. All regressions include country, industry and time dummies. Heteroscedasticity-consistent robust standard errors are clustered at the country. * Significant at 10%; ** significant at 5%; *** significant at 1% 39   Table 9: Sub-sovereign ratings below sovereign ratings controlling for whether sub- sovereign debt is issued (Heckman selection model) S&P Moody’s Fitch (1) (2) (3) Coeff. Std. Err. Coeff. Std. Err. Coeff. Std. Err. (Average marginal effect) A. Prob. below ceiling Public entity -0.404*** (0.026) -0.474*** (0.028) -0.354*** (0.035) Financial sector -0.125*** (0.025) -0.055** (0.027) 0.152*** (0.034) GDP growth 0.014*** (0.005) 0.013** (0.006) 0.026*** (0.009) Private credit/GDP 0.007*** (0.001) 0.011*** (0.001) 0.007*** (0.003) Inflation rate -0.005** (0.002) -0.001 (0.001) -0.001 (0.002) Cur. Ac. Bal./GDP -0.019*** (0.005) -0.019*** (0.005) -0.016* (0.009) Trade/GDP 0.002* (0.001) 0.000 (0.001) 0.002 (0.002) Cap. ac. Openness 0.177** (0.087) 0.404*** (0.096) -0.199 (0.192) US 3-month int. rate 0.011 (0.038) 0.048 (0.043) 0.078 (0.055) US 10-yr bond yield -0.125*** (0.039) -0.186*** (0.041) -0.167*** (0.053) VIX index -0.045*** (0.004) -0.051*** (0.004) -0.041*** (0.005) GFC dummy 0.077 (0.097) 0.139 (0.107) 0.215** (0.107) Constant -2.255*** (0.422) -1.881*** (0.336) -3.544*** (0.496) Excluded variable Governed by English common law 0.594*** (0.027) 0.595*** (0.029) 0.623*** (0.036) B. Distance below ceiling Public entity -1.746*** (0.117) -2.063*** (0.133) -2.383*** (0.132) Financial sector -0.576*** (0.097) -1.171*** (0.101) -0.206* (0.121) GDP growth 0.018 (0.023) -0.006 (0.024) -0.004 (0.034) Private credit/GDP -0.018*** (0.004) 0.010 (0.006) 0.006 (0.010) Inflation rate -0.022** (0.010) -0.010 (0.012) 0.055* (0.030) Cur. Ac. Bal./GDP -0.065*** (0.021) 0.029 (0.023) 0.065** (0.031) Trade/GDP -0.014*** (0.005) -0.017*** (0.005) -0.008 (0.008) Cap. ac. Openness 0.841*** (0.320) 0.456 (0.375) -1.240** (0.630) US 3-month int. rate 0.155 (0.145) 0.055 (0.172) 0.243 (0.197) US 10-yr bond yield -0.014 (0.138) 0.077 (0.145) 0.094 (0.161) VIX index -0.029* (0.016) -0.055*** (0.018) -0.030* (0.017) GFC dummy 0.190 (0.348) -0.289 (0.385) 0.088 (0.350) Constant 1.864 (1.844) -1.018 (1.299) -2.032 (2.046) Lambda (Inverse Mills ratio) 0.125 (0.206) 0.373* (0.216) 0.642*** (0.234) Country & year dummies Y Y Y Chi2 1227.4 572.32 818.45 Prob > Chi2 0.000 0.000 0.000 No. of obs. 43,905 43,631 43,073 Notes: The regression is estimated using a two-step Heckman procedure. Ratings are on a scale of 1-21, with higher values indicating better ratings. See table 1 for conversion from letter to numeric scale. Debt in default and SPVs are excluded. All regressions include country and year dummies. * Significant at 10%; ** significant at 5%; *** significant at 1% 40   Table 10: Sub-sovereign ratings above sovereign ratings controlling for whether sub- sovereign debt is issued (Heckman selection model) S&P Moody’s Fitch (1) (2) (3) Coeff. Std. Err. Coeff. Std. Err. Coeff. Std. Err. (Average marginal effect) A. Prob. above ceiling Public entity -0.448*** (0.025) -0.504*** (0.026) -0.437*** (0.034) Financial sector -0.124*** (0.024) 0.031 (0.025) 0.190*** (0.033) GDP growth 0.014*** (0.005) 0.016*** (0.005) 0.024*** (0.009) Private credit/GDP 0.008*** (0.001) 0.011*** (0.001) 0.006*** (0.002) Inflation rate -0.005*** (0.002) -0.002 (0.001) -0.001 (0.002) Cur. Ac. Bal./GDP -0.016*** (0.005) -0.020*** (0.005) -0.017** (0.008) Trade/GDP 0.002* (0.001) 0.001 (0.001) 0.002 (0.002) Cap. ac. Openness 0.224*** (0.081) 0.335*** (0.087) -0.103 (0.183) US 3-month int. rate -0.014 (0.036) 0.01 (0.036) 0.109** (0.053) US 10-yr bond yield -0.129*** (0.037) -0.186*** (0.037) -0.177*** (0.051) VIX index -0.045*** (0.004) -0.047*** (0.004) -0.041*** (0.005) GFC dummy 0.067 (0.095) 0.176** (0.089) 0.206** (0.102) Constant -2.005*** (0.400) -1.653*** (0.305) -3.513*** (0.479) Excluded variable Governed by English common law 0.583*** (0.026) 0.583*** (0.026) 0.619*** (0.035) B. Distance above ceiling Public entity -0.525*** (0.096) -0.351*** (0.108) -0.257*** (0.074) Financial sector 0.316*** (0.077) 0.554*** (0.080) 0.201*** (0.064) GDP growth 0.018 (0.017) -0.006 (0.019) -0.042** (0.018) Private credit/GDP 0.003 (0.003) -0.004 (0.005) -0.013*** (0.005) Inflation rate -0.003 (0.007) 0.017** (0.009) 0.02 (0.016) Cur. Ac. Bal./GDP 0.003 (0.016) -0.007 (0.018) -0.026 (0.016) Trade/GDP 0.004 (0.004) 0.006 (0.004) 0.008* (0.004) Cap. ac. Openness 0.410* (0.246) -0.075 (0.286) 0.391 (0.328) US 3-month int. rate -0.283*** (0.109) -0.343*** (0.120) 0.174* (0.102) US 10-yr bond yield -0.053 (0.108) -0.001 (0.116) 0.027 (0.085) VIX index -0.005 (0.012) 0.002 (0.013) 0.003 (0.009) GFC dummy 0.022 (0.285) -0.2 (0.279) 0.109 (0.181) Constant 0.925 (1.475) 0.527 (1.046) 0.003 (1.070) Lambda (Inverse Mills ratio) 0.425*** (0.165) 0.464*** (0.177) 0.148 (0.124) Country & year dummies Y Y Y Chi2 431.2 572.3 238.2 Prob > Chi2 0.000 0.000 0.000 No. of obs. 43,884 43,818 42,912 Notes: The regression is estimated using a two-step Heckman procedure. Ratings are on a scale of 1-21, with higher values indicating better ratings. See table 1 for conversion from letter to numeric scale. Debt in default and SPVs are excluded. All regressions include country and year dummies. * Significant at 10%; ** significant at 5%; *** significant at 1% 41   Figure 1a: Relationship between Sovereign and Sub-sovereign debt ratings for Standard & Poor’s S&P: Public Entity S&P: Private Entity AA+ AA+ Sub-sovereign Rating Sub-sovereign Rating A- A- BB BB CCC+ CCC+ CCC+ BB A- AA+ CCC+ BB A- AA+ Sovereign Rating Sovereign Rating S&P: Non-Financial issuers S&P: Financial issuers AA+ AA+ Sub-sovereign Rating Sub-sovereign Rating A- A- BB BB CCC+ CCC+ CCC+ BB A- AA+ CCC+ BB A- AA+ Sovereign Rating Sovereign Rating   42   Figure 1b: Relationship between Sovereign and Sub-sovereign debt ratings for Moody’s Moody's: Public Entity Moody's: Private Firms AA+ AA+ Sub-sovereign Rating Sub-sovereign Rating A- A- BB BB CCC+ CCC+ CCC+ BB A- AA+ CCC+ BB A- AA+ Sovereign Rating Sovereign Rating Moody's: Non-Financial Issuers Moody's: Financial Issuers AA+ AA+ A- A- Sub-sovereign Rating Sub-sovereign Rating BB BB CCC+ CCC+ CCC+ BB A- AA+ CCC+ BB A- AA+ Sovereign Rating Sovereign Rating   43   Figure 1c: Relationship between Sovereign and Sub-sovereign debt ratings for Fitch Fitch: Public Entity Fitch: Private Firms AA+ AA+ A- A- Sub-sovereign Rating Sub-sovereign Rating BB BB CCC+ CCC+ CCC+ BB A- AA+ CCC+ BB A- AA+ Sovereign Rating Sovereign Rating Fitch: Non-Financial Issuers Fitch: Financial Issuers AA+ AA+ A- A- Sub-sovereign Rating Sub-sovereign Rating BB BB CCC+ CCC+ CCC+ BB A- AA+ CCC+ BB A- AA+ Sovereign Rating Sovereign Rating       44   Appendix Tables Table A.1: Sub-sovereign debt ratings below the sovereign rating: Results of the Tobit model for non-financial issuers S&P Moody’s Fitch (1) (2) (3) Coeff. Std. Err. Coeff. Std. Err. Coeff. Std. Err. Average marginal effects Distance below ceiling -0.527*** (0.118) -0.474*** (0.136) -0.532*** (0.129) Log Issue size -0.034*** (0.013) -0.064*** (0.011) -0.056*** (0.014) Maturity (years) 0.803** (0.325) 0.279 (0.402) -0.204 (0.713) Fixed rate note 1.072*** (0.234) 1.200*** (0.248) 0.770*** (0.186) Callable bond 0.271 (0.247) 0.291 (0.215) 0.500 (0.313) Collateralized -0.042 (0.236) -0.073 (0.345) 0.668* (0.350) Euro issue -1.107*** (0.365) -1.979*** (0.481) -0.944* (0.571) Yen issue -0.913*** (0.251) -1.463*** (0.380) -1.806*** (0.261) Public entity 0.802** (0.316) 0.355 (0.352) 0.379 (0.248) Non-tradable -1.572** (0.773) -1.308** (0.615) -1.879* (0.979) Local Gov. -0.051 (0.055) -0.018 (0.045) -0.045 (0.050) GDP growth -0.012 (0.011) 0.001 (0.010) -0.001 (0.009) Private credit/GDP -0.038 (0.027) -0.019 (0.022) 0.025 (0.061) Inflation rate -0.075* (0.041) 0.019 (0.046) 0.084* (0.044) Cur. ac. bal./GDP -0.013 (0.016) -0.013 (0.009) -0.004 (0.009) Trade/GDP 1.394* (0.829) 1.445** (0.656) 0.110 (1.138) Cap. ac. openness 0.231 (0.282) -0.14 (0.257) -0.249 (0.312) US 3-month int. rate -0.011 (0.132) -0.004 (0.159) 0.242 (0.247) US 10-yr bond yield -0.032 (0.020) -0.046*** (0.014) -0.008 (0.026) VIX index 1.707*** (0.057) 1.630*** (0.054) 1.557*** (0.171) Constant Y Y Y Country & Ind. dummies 0.190 0.257 0.167 Pseudo R-square -2761.9 -2073.6 -1227.2 Log likelihood 1,584 (37) 1,279 (36) 710 (30) No. of obs. (countries) Notes: The dependent variable is the distance of the sub-sovereign debt rating below the sovereign rating. All regressions include country, industry and time dummies. Heteroscedasticity-consistent robust standard errors are clustered at the country level            Table A.2: Sub-sovereign debt ratings below the sovereign rating: Results of the Tobit model for financial issuers S&P Moody’s Fitch (1) (2) (3) Coeff. Std. Err. Coeff. Std. Err. Coeff. Std. Err. Average marginal effects Distance below ceiling -0.021 (0.063) -0.088 (0.071) -0.142 (0.111) Log Issue size 0.012 (0.009) 0.016*** (0.006) 0.001 (0.006) Maturity (years) 0.176 (0.166) 0.427** (0.194) 0.413* (0.244) Fixed rate note 0.434*** (0.132) 0.256** (0.114) 0.504*** (0.123) Callable bond -0.516*** (0.160) -0.454*** (0.155) -0.521** (0.229) Collateralized -0.057 (0.123) -0.018 (0.167) 0.326* (0.181) Euro issue 0.101** (0.048) 0.055 (0.079) -0.719*** (0.142) Yen issue -1.105*** (0.229) -1.000*** (0.177) -1.443*** (0.301) Public entity 0.395*** (0.149) 0.248*** (0.082) 0.487*** (0.180) Bank 0.018 (0.026) -0.018 (0.031) 0.015 (0.026) GDP growth -0.015*** (0.005) -0.028*** (0.007) -0.004 (0.005) Private credit/GDP -0.016* (0.009) -0.027 (0.031) 0.046 (0.034) Inflation rate 0.036 (0.027) -0.007 (0.019) 0.006 (0.019) Cur. ac. bal./GDP -0.007 (0.007) -0.004 (0.009) -0.001 (0.004) Trade/GDP 0.489 (0.455) 0.900* (0.469) 0.110 (0.455) Cap. ac. openness 0.201*** (0.041) 0.021 (0.055) 0.219** (0.101) US 3-month int. rate 0.078 (0.069) 0.057 (0.105) -0.086 (0.107) US 10-yr bond yield -0.013*** (0.005) -0.036*** (0.007) -0.027*** (0.010) VIX index -0.777*** (0.232) -1.204*** (0.125) -0.536*** (0.208) Constant Y Y Y Country & Ind. dummies 0.235 0.268 0.259 Pseudo R-square -2014.8 -1565.6 -1373.2 Log likelihood 1,847 (39) 1,619 (39) 1,215 (45) No. of obs. (countries) Notes: The dependent variable is the distance of the sub-sovereign debt rating below the sovereign rating. All regressions include country, industry and time dummies. Heteroscedasticity-consistent robust standard errors are clustered at the country level    46   Table A.3: Sub-sovereign debt ratings above the sovereign rating: Results of the Double-Hurdle model for non-financial issuers S&P Moody’s Fitch (1) (2) (3) Coeff. Std. Err. Coeff. Std. Err. Coeff. Std. Err. Average marginal effects I(above ceiling) 0.012 (0.014) 0.048*** (0.018) 0.027* (0.015) Log Issue size 0.001 (0.001) 0.002* (0.001) 0.001 (0.001) Maturity (years) 0.287*** (0.050) 0.420*** (0.145) 0.240*** (0.056) Future-flow receivable 0.078** (0.033) -0.024 (0.037) 0.050 (0.036) Other SPV 0.048 (0.039) 0.041 (0.066) -0.033 (0.026) Other collateralized -0.034 (0.031) -0.080** (0.032) -0.093*** (0.036) Fixed rate note -0.027 (0.045) -0.031 (0.040) -0.008 (0.029) Callable bond 0.036 (0.024) 0.079** (0.033) -0.028 (0.027) Euro issue 0.104 (0.075) 0.230** (0.102) 0.004 (0.084) Yen issue -0.023 (0.018) -0.002 (0.060) -0.069* (0.037) Public entity -0.073* (0.043) -0.017 (0.043) -0.094*** (0.030) Non-tradable -0.205** (0.090) -0.200*** (0.068) … … Local gov. 0.006 (0.004) 0.019*** (0.005) -0.002 (0.007) GDP growth 0.000 (0.001) -0.001 (0.001) -0.002** (0.001) Private credit/GDP 0.000 (0.002) 0.005 (0.003) -0.003 (0.005) Inflation rate 0.010*** (0.003) 0.005 (0.004) 0.013*** (0.005) Cur. ac. bal./GDP -0.004*** (0.001) -0.005*** (0.001) 0.001 (0.001) Trade/GDP 0.083 (0.064) -0.074 (0.083) -0.119 (0.169) Cap. ac. openness -0.030 (0.024) -0.056 (0.037) -0.011 (0.023) US 3-month int. rate -0.013 (0.015) 0.005 (0.030) 0.009 (0.029) US 10-yr bond yield 0.005*** (0.001) 0.009*** (0.002) 0.002* (0.001) VIX index 0.150*** (0.003) 0.196*** (0.006) 0.113*** (0.006) Constant Distance above ceiling -0.024 (0.048) -0.014 (0.065) 0.007 (0.046) Log Issue size 0.007*** (0.003) 0.002 (0.003) 0.003 (0.003) Maturity (years) 1.087*** (0.140) 1.595*** (0.326) 0.864*** (0.095) Future-flow receivable 0.340** (0.147) 0.194 (0.164) 0.285*** (0.090) Other SPV 0.224* (0.129) 0.365** (0.155) -0.172 (0.135) Other collateralized -0.049 (0.097) -0.020 (0.057) -0.220 (0.155) Fixed rate note -0.042 (0.113) -0.062 (0.104) -0.016 (0.103) Callable bond -0.043 (0.069) 0.065 (0.100) -0.022 (0.049) Euro issue 0.297 (0.241) 0.606* (0.343) -0.060 (0.182) Yen issue -0.188** (0.091) -0.008 (0.138) 0.213* (0.119) Public entity -0.104 (0.135) -0.082 (0.116) -0.229** (0.113) Non-tradable -0.887** (0.356) -0.786 (0.501) … … Local gov. 0.007 (0.014) 0.036** (0.014) -0.002 (0.022) GDP growth 0.001 (0.003) 0.000 (0.004) -0.004 (0.003) Private credit/GDP -0.006 (0.005) 0.007 (0.009) 0.016 (0.015) Inflation rate 0.034*** (0.012) 0.020* (0.011) 0.032** (0.015) Cur. ac. bal./GDP -0.014*** (0.003) -0.020*** (0.004) -0.001 (0.004) Trade/GDP 47   0.204 (0.214) -0.155 (0.203) -0.031 (0.526) Cap. ac. openness -0.122 (0.082) -0.175* (0.094) -0.044 (0.042) US 3-month int. rate 0.045 (0.052) 0.114 (0.077) -0.030 (0.083) US 10-yr bond yield 0.015** (0.006) 0.024*** (0.007) 0.009*** (0.003) VIX index 0.467*** (0.027) 0.607*** (0.040) 0.260*** (0.036) Constant Y Y Y Country & Ind. dummies 0.270 0.274 0.293 Pseudo R-square -1009.5 -1054.7 -312.6 Log likelihood 1,863 (37) 1,590 (37) 800 (30) No. of obs. (countries) Notes: The dependent variable in the first block is an indicator that the sub-sovereign debt rating is higher than the sovereign rating, and dependent variable in the second block the distance between the two ratings. All regressions include country, industry and time dummies. Heteroscedasticity-consistent robust standard errors are clustered at the country level. 48   Table A.4: Sub-sovereign debt ratings above the sovereign rating: Results of the Double-Hurdle model for financial issuers S&P Moody’s Fitch (1) (2) (3) Coeff. Std. Err. Coeff. Std. Err. Coeff. Std. Err. Average marginal effects I(above ceiling) 0.008 (0.006) 0.029* (0.017) -0.004 (0.007) Log Issue size 0.001 (0.001) 0.003 (0.002) 0.000 (0.002) Maturity (years) 0.184*** (0.023) 0.411*** (0.087) 0.289*** (0.069) Future-flow receivable 0.080*** (0.018) 0.155 (0.110) 0.136*** (0.047) Other SPV 0.077** (0.031) 0.338*** (0.056) … … Other collateralized -0.005 (0.010) -0.023 (0.020) 0.004 (0.017) Fixed rate note 0.004 (0.016) -0.026 (0.031) 0.022 (0.025) Callable bond 0.001 (0.017) -0.075* (0.042) 0.009 (0.051) Euro issue -0.036*** (0.007) 0.009 (0.029) … … Yen issue -0.009 (0.011) 0.077 (0.074) -0.088* (0.050) Public entity -0.023** (0.010) 0.063 (0.078) -0.002 (0.021) Bank -0.002 (0.002) -0.009 (0.007) -0.006 (0.004) GDP growth 0.000 (0.000) 0.000 (0.001) 0.001* (0.001) Private credit/GDP -0.001*** (0.000) 0.002 (0.007) 0.008 (0.005) Inflation rate -0.002* (0.001) -0.004 (0.007) -0.013** (0.005) Cur. ac. bal./GDP 0.000 (0.000) 0.000 (0.001) 0.000 (0.001) Trade/GDP -0.033 (0.034) -0.139** (0.065) 0.049 (0.082) Cap. ac. openness -0.017** (0.009) -0.047* (0.026) 0.005 (0.015) US 3-month int. rate 0.009 (0.010) 0.044* (0.025) -0.002 (0.018) US 10-yr bond yield 0.000 (0.001) 0.014*** (0.002) -0.003 (0.003) VIX index 0.099*** (0.002) 0.299*** (0.005) 0.118*** (0.014) Constant Distance above ceiling 0.062** (0.026) 0.109*** (0.042) -0.008 (0.020) Log Issue size 0.008* (0.005) -0.005 (0.007) -0.005 (0.005) Maturity (years) 0.963*** (0.094) 1.802*** (0.164) 0.845*** (0.177) Future-flow receivable 0.634*** (0.072) 1.229*** (0.391) 0.592*** (0.127) Other SPV 0.239 (0.154) 0.918*** (0.167) … … Other collateralized -0.232*** (0.043) -0.097 (0.073) 0.019 (0.048) Fixed rate note -0.030 (0.082) -0.246** (0.099) -0.032 (0.057) Callable bond -0.122 (0.111) -0.212** (0.089) 0.026 (0.120) Euro issue -0.240 (0.157) 0.091 (0.071) … … Yen issue -0.440** (0.218) 0.225 (0.160) -0.251 (0.182) Public entity 0.037 (0.080) -0.029 (0.226) -0.004 (0.107) Bank -0.021 (0.018) -0.077*** (0.023) -0.027** (0.011) GDP growth -0.002* (0.001) -0.003 (0.002) 0.002 (0.001) Private credit/GDP 0.001 (0.002) 0.004 (0.013) 0.014 (0.008) Inflation rate -0.007 (0.013) -0.018 (0.020) -0.016 (0.011) Cur. ac. bal./GDP 49   0.000 (0.001) -0.002 (0.002) 0.000 (0.001) Trade/GDP -0.326** (0.141) -0.259 (0.216) 0.067 (0.145) Cap. ac. openness -0.091* (0.050) -0.117 (0.078) 0.003 (0.026) US 3-month int. rate 0.129 (0.085) 0.383*** (0.082) 0.045 (0.071) US 10-yr bond yield 0.001 (0.005) 0.034*** (0.007) -0.012 (0.007) VIX index 0.519*** (0.022) 0.905*** (0.068) 0.294*** (0.032) Constant Y Y Y Country & Ind. dummies 0.422 0.227 0.341 Pseudo R-square -693.3 -2216.0 -496.9 Log likelihood 2,051 (41) 2,318 (44) 1,293 (37) No. of obs. (countries) Notes: The dependent variable in the first block is an indicator that the sub-sovereign debt rating is higher than the sovereign rating, and dependent variable in the second block the distance between the two ratings. All regressions include country, industry and time dummies. Heteroscedasticity-consistent robust standard errors are clustered at the country level. 50