گسترش وام و اعتبار در اوراق قرضه شرکتی و اقتصاد کلان در ژاپن

Using secondary market data on corporate bonds issued in Japan between 1997 and 2005, this paper explores the determinants of the credit spread of corporate bond rates over interest swap rates. We find that the credit spreads properly reflect financial factors at the firm level, including debt-to-equity ratios, volatility, and maturity, particularly for longer-term bonds. In addition, an economy-wide factor common among bond issues unable to be captured by firm-level factors, plays an important role in determining credit spreads, and these economy-wide effects to a great extent cancel out firm-level factors for some subsample periods. We also identify possible factors responsible for the significant economy-wide effects.

Using secondary market data on corporate bonds issued in Japan between 1997 and 2005, this paper empirically investigates the possible determinants of credit spreads on corporate bonds, including financial factors summarized at the level of individual firms as well as macroeconomic and market-wide effects. According to the standard frameworks for bond pricing models, including Merton (1974), credit risks mainly reflect firm-level financial factors responsible for the possibility of individual default, while interest rate risks are only determined by market-wide factors common among individual firms. Typically, these include macroeconomic conditions and monetary policies. One of the more important implications of this model is that firm-specific and macroeconomic factors responsible for the determination of credit spreads may be summarized by variables at the individual firm level. The risk-free rate is the only macroeconomic variable that appears in the standard model. Given this conventional prediction, as long as the set of firm-level explanatory variables is properly chosen to reflect both the firm-specific and macroeconomic components, the credit spreads of corporate bond rates over market interest rates can be explained mostly by firm-level financial conditions. These include debt-to-equity ratios and the volatility of corporate value, along with individual contract clauses, such as maturity and any attached options. In other words, there is little room for credit spreads to be influenced by market-wide effects (except for the risk-free rate) beyond what is captured by these firm-level financial variables and contract clauses. We carefully and rigorously assess the empirical relevance of the above prediction by raising the following questions: namely, (i) whether credit spreads on corporate bonds reflect firm-level financial factors in a proper way; (ii) whether there are market-wide effects other than firm-level factors; and (iii) if the answer to the second question is in the affirmative, what macroeconomic conditions are responsible for market-wide factors. Our empirical investigation is motivated mainly by the following observation concerning Japanese corporate bond markets. One of the clear and simple predictions available from standard pricing models of credit risk is the negative correlation between credit spreads and equity prices, which serves as a proxy for corporate valuation. That is, a decrease in equity prices will enhance default risk and thereby raise credit spreads on corporate bonds. Fig. 1 plots the relation between the average credit spreads on Moody’s A-rated corporate bonds and the average total equity valuation of the issuing firms. As shown, there is indeed a negative correlation between credit spreads and equity valuations for both the period between 1997 and 2002, and the period between 2003 and 2005. Between 2002 and 2003, however, credit spreads declined substantially although equity valuations also fell heavily. The positive correlation in these subsample periods is uniformly observed for highly rated corporate bonds with different maturities (from less than three years to longer than 10 years). Among low-grade corporate bonds, such as Baa rated issues, a positive correlation between credit spreads and equity valuation is again observed for the period between 2001 and 2003 (see Fig. 2). In addition, a positive correlation between the two is observed between 1997 and 1999. Credit spreads increased while equity valuation was relatively strong. Such overall consistency and particular inconsistency in the relationship between credit spreads and corporate valuation may help to separate independent market-wide effects on credit spreads from firm-specific factors.The motivation of this paper is shared with the existing empirical literature on US corporate bond pricing. Among the empirical papers based on corporate bonds issued in the US, Collin-Dufresne et al., 2001 and Delianedis and Geske, 2001 divided the determinants of credit spreads into market-wide factors and firm-level factors. Collin-Dufresne et al. (2001) found that firm-level financial factors, including leverage ratios and equity valuation, play little part in determining credit spreads and that credit spreads are largely subject to market-wide factors possibly associated with overall market liquidity. Delianedis and Geske (2001) established that the firm-level financial factors, including the volatility of corporate value, did not contribute to the determination of credit spreads on corporate bonds and that individual credit spreads were heavily influenced by market risks measured in terms of the returns and volatilities of equity market indexes. In addition, Jones et al., 1984 and Huang and Huang, 2003 demonstrated that firm-level financial factors do not contribute to corporate bond pricing.1 Together, these papers suggest that the determination of credit spreads is seriously inconsistent with the standard bond pricing theory; in fact, these empirical results act partly as a trigger for the theoretical development of more general models. In terms of empirical studies concerning corporate bonds issued in Japan, Ueki, 1999, Ieda and Ohba, 1998 and Ieda, 2001 examined the possible determinants of credit spreads, claiming that firm-specific factors are mainly responsible for the determination of credit spreads.2 However, the fundamental difference from the work based on US corporate bonds and our analysis is that they investigated the relationship between credit spreads and credit ratings, and was unconcerned with the possible effects of the firm-level financial factors underlying these ratings. Our major findings are summarized as follows. First, credit spreads properly reflect firm-level financial factors, including equity valuation and the volatility of corporate value, particularly for corporate bonds with maturities in excess of 10 years. Second, economy-wide effects also play an important role in determining credit spreads. For the period between 1997 and 1999, and again between 2001 and 2003, an economy-wide effect dominated and canceled out the effects dictated by the firm-level financial conditions, thereby yielding a positive correlation between credit spreads and equity valuation at the aggregate level. Third, further empirical investigation into market-wide effects demonstrates that the overall deterioration of corporate bond market liquidity during a financial crisis contributed to a significantly positive market-wide effect on credit spreads from 1997 through 1999. Moreover, massive capital inflows into corporate bond markets because of aggressive monetary policy generated a significant aggregate impact between 2001 and 2003. The remainder of the paper is organized as follows. Section 2 reviews the empirical predictions based on the standard credit risk model. Sections 3 and 4 present the empirical specifications and estimation results for firm-level (or issue-specific) effects and market-wide effects, respectively. Section 5 offers some conclusions.

Using data on corporate bonds issued in Japan between 1997 and 2005, this paper considers the possible determinants of the credit spreads of corporate bond rates over interest swap rates. We find that the credit spreads reasonably reflect firm-level financial factors, including debt-to-equity ratios, volatility, and maturity. Overall, the results indicate that firm and issue-specific factors influence credit spreads in a quite reasonable manner. These findings contrast sharply with a similar work on US corporate bonds where firm level financial conditions were found not to play any significant role in determining individualcredit spreads. In this regard, corporate bond pricing in the Japanese market is more consistent with a standard version of Merton (1974) than the US market. On the other hand, an economy-wide factor common among bond issues, as measured by time ef- fects, plays an important role in determining credit spreads. This aspect is seriously inconsistent with Merton’s (1974) standard model where macroeconomic effects are mostly captured by firm-level vari- ables along with risk-free rates. That is, the Japanese market shares with the US market the feature that market-wide effects, including market liquidity, are significant determinants of credit spreads. This common factor had particularly significant effects on the credit spreads observed between 1997 and 1998, when financial markets were subject to liquidity crises, and between 2001 and 2003, when the Bank of Japan implemented a quantity-easing policy with zero overnight money mar- ket rates. During both the periods, the economy-wide effect largely canceled out the firm-level factors. In the earlier period, credit spreads increased even though individual stock prices (or equivalently cor- porate values) were still firm, while in the more recent period, credit spreads declined substantially, although equity valuation also fell heavily. The empirical analysis of the more recent period indicates that credit risks valuated downwards because of the rich liquidity in corporate bond markets. One limitation of our empirical analysis is that we ignore issue-specific or firm-specific liquidity factors by assuming that liquidity effects are market wide. Among the recent work in this area, Chen et al. (2007) demonstrate empirically that US credit spreads are subject not only to market-wide liquidity factors but also to issue-specific liquidity, as measured by the issue-by-issue bid–ask spreads and the frequency of individual transactions. Ericsson and Renault (2006) present theoretically the interaction between credit risks and issue-specific liquidity. We would like to extend our research along this line of inquiry