Chapter

Chapter 4 Credit Bonds

Editor(s):
Alfred Schipke, Markus Rodlauer, and Longmei Zhang
Published Date:
March 2019
Share
  • ShareShare
Show Summary Details
Author(s)
ZHANG Longmei and WU Yuchen 

The authors thank CHEN Wanhong of the National Development and Reform Commission and LU Dabiao of the China Securities Regulatory Commission for helpful comments.

China’s credit bond market has grown rapidly in recent decades. Since 1983, when China’s first enterprise bond was issued, the credit bond market has expanded from enterprise bonds to new products such as corporate bonds and medium-term notes, among others. By the end of 2017, outstanding credit bonds reached $2.7 trillion, the second-largest market in the world after the United States (Figure 4.1).

Figure 4.1.Outstanding Debt Securities of Nonfinancial Corporations

(Trillions of US dollars)

Sources: Bank for International Settlements 2017.

Note: Data are as of the third quarter of 2017.

Bond financing remains a small portion of corporate financing in China, however, suggesting that there is still significant room for expansion (Figure 4.2). Credit bonds account for only 30 percent of GDP and 10 percent of total financing of nonfinancial corporations. Banks continue to dominate, with bank loans accounting for 70 percent of corporate financing, compared to around 10 percent in the United States. As the financial system moves toward direct financing, credit bonds are likely to continue growing rapidly in the long term.

Figure 4.2.Nonfinancial Corporation Financing, by Type, 2017

Source: People’s Bank of China 2017.

Foreign participation remains extremely low, accounting for only 1 percent of the overall market, compared to 29 percent in the United States. Recent measures to open up the bond market, such as granting institutional investors direct access to the interbank market and Bond Connect,1 have mostly attracted foreign inflows to the sovereign bond market rather than to credit bonds. Significant room for increased foreign participation remains.

Despite the huge potential, China’s credit bond market is beset by several issues that will be important to overcome. First, the market is highly segmented, with different bonds subject to different regulatory requirements and traded and settled on different platforms. Second, reflecting market-perceived government guarantees, especially for state-owned enterprises (SOEs) and local government financing vehicles (LGFVs), credit ratings appear to be distorted—with more than 90 percent of credit bonds rated above AA. While the government has recently issued a series of new regulations to break such implicit guarantees, in practice they still exist. Hence, bond pricings may not adequately reflect underlying risks, leading to inefficient credit allocation and large contingent government liabilities. Access to the credit bond market remains quite restrictive, with issuance dominated by SOEs, while private firms have very limited access.

This chapter studies these issues in detail, in particular, how the current credit bond market is structured, how liquidity risks and credit risks are priced for different credit bonds, and what role implicit government guarantees play in bond pricing. The rest of the chapter first discusses the regulatory structure and characteristics of the various credit bond segments then analyzes empirically the pricing behavior of different types of credit bonds, and tests the impact of implicit guarantees on bond pricing.

Market Structure and Characteristics

China’s credit bond market is highly diverse, with many different market segments.2 Enterprise bonds, corporate bonds, and medium-term notes and commercial paper account for the lion’s share (Figure 4.3). Nonfinancial firms can in principle freely choose any type of bond as a financing tool, but for historical reasons, SOEs typically dominate enterprise bond issuance, while private firms normally issue corporate bonds or medium-term notes or commercial paper.

Figure 4.3.Composition of the Credit Bond Market, 2017

Source: WIND Economic Database (www.wind.com.cn) 2017.

Credit bond regulation differs across agencies. Unlike in the United States, where the Securities and Exchange Commission is the sole regulator, credit bonds in China are regulated by different regulatory agencies, subject to different issuance procedures, and traded and settled on different platforms (Figure 4.4). Among the three main types of credit bonds, enterprise bonds are regulated by the National Development and Reform Commission according to the Corporate Law, the Securities Law, and Guidance on Enterprise Bond Management, and traded on both the interbank market and the exchange market. Corporate bonds are regulated by the China Securities Regulatory Commission according to the Securities Law and the Guideline on Corporate Issuance and Trading, and traded only on the exchange market. Medium-term notes and commercial paper are regulated by the National Association of Financial Market Institutional Investors under the People’s Bank of China, based on the Guideline on Interbank Bond Market Nonfinancial Firms Financial Tools, and traded only on the interbank market.

Figure 4.4.Segmented Bond Market Regulation

Source: Authors’ compilation.

Note: CCDC = China Central Depository and Clearing Corporation Limited; CSDC = China Securities Depository and Clearing Corporation Limited; CSRC = China Securities Regulatory Commission; NAFMII = National Association of Financial Market Institutional Investors; NDRC = National Development and Reform Commission.

In terms of issuance procedures, medium-term notes and commercial paper are registration based, while the issuance of enterprise bonds still requires approval. For corporate bonds, issuance to the general public requires approval, while private placement is registration based. All credit bonds (excluding private placement) require minimum ratings, such as AA or A–, or in some cases AAA. For firms with a leverage ratio exceeding a certain threshold, a compulsory external guarantee is required, in addition to the minimum ratings requirement.3

The existence of similar credit bond markets reflects China’s unique historical development of credit bonds (Figure 4.5), which is closely related to the country’s broader reform and policies to open its markets.

Figure 4.5.Development of Credit Bond, 1983–2017

Source: Authors’ compilation.

1 Commercial paper was suspended in 1997.

The development of China’s credit bond market has proceeded in three stages:

  • First stage—enterprise bonds: For a long period, enterprise bonds were the only type of credit bonds in China. They were first issued in 1983 and corresponding formal regulation was introduced in 1987 by the State Council. Overall, the guiding principle for enterprise bonds is to channel low-cost funding to weak links in the economy and achieve the goals of the national development strategy. Historically, enterprise bonds have been designed to provide financing for key national infrastructure projects, such as the Three Gorges Hydraulic Project, the national railway construction project, the construction and reengineering of the national electrical grid, and the development of the steel, chemical engineering, energy, and other industries. Over time, as China’s development strategy has evolved, enterprise bonds have also expanded to finance such areas as green development, social housing, poverty reduction, and social services for seniors.

    Initially, enterprise bonds were jointly regulated by the People’s Bank of China and the National Development and Reform Commission (NDRC). At the end of the 1990s, purview shifted solely to the NDRC. Each bond issuance is typically earmarked for a project. The NDRC is in charge of both project approval and monitoring, and the corresponding bond issuance. With the reform in recent years, firms now have more flexibility in the bond issuance schedule within the quota that has been granted. For historical reasons, enterprise bond issuance is almost entirely by SOEs, both at the central and local levels. Bonds issued by LGFVs (so-called chengtou zhai), now account for 80 percent of enterprise bonds. By 2017, enterprise bonds accounted for 17 percent of outstanding credit bonds.

  • Second stage—corporate bonds: Since 2004, the pace of capital market liberation has quickened, which has led to the introduction of new financial products, including bonds issued by financial institutions (so-called financial bonds) and asset-backed securities. In 2007, corporate bonds were launched, regulated by the China Securities Regulatory Commission and traded solely on the exchange market. Initially, the issuance was restricted to listed companies. Later, the regulator streamlined approval and filing procedures for exchange membership, and established a private placement system for nonlisted companies in 2015, which has spurred the rapid growth of this segment of the market. The panda bond, a renminbi-denominated bond issued by offshore companies, was also introduced to the exchange bond market in 2016. By 2017, outstanding corporate bonds were worth $600 billion, accounting for 25 percent of the entire credit bond market.

  • Third stage—medium-term notes and commercial paper: To further promote the ability of companies to raise financing directly, commercial paper was reintroduced in 2005, and a new type of credit bond, medium-term notes, was launched in 2008. Both commercial paper and medium-term notes are formally classified as nonfinancial-firm debt financing instruments,4 regulated by the National Association of Financial Market Institutional Investors (which is legally part of the People’s Bank of China) and traded on the interbank market. Medium-term notes are the first type of credit bonds that can be issued on a registration basis, instead of after formal approval by the respective regulators. By 2017, the outstanding stock was $600 billion, accounting for another 26 percent of total credit bonds outstanding. Commercial paper with tenors of 3, 6, 9, or 12 months reached $50 billion, about 8 percent of total credit bonds.

Trading

The trading platform, meanwhile, has shifted from the exchange market to the interbank bond market. In 1997, about 98 percent of credit bonds were traded on the exchange market. This percentage declined significantly to 7 percent by 2017, while the remaining 93 percent of transactions are traded in the interbank market (Figure 4.6).5

Figure 4.6.Credit Bond Trading, by Market, 2017

Source: WIND Economic Database (www.wind.com.cn) 2017.

The current trading structure is in line with international experience, which shows that 90 percent of bond transactions typically take place outside the exchange market, that is, over the counter. It is notable that over-the-counter bond transactions in China take place in the interbank bond market, which typically provides funding only for banks in other countries. The interbank market is a wholesale market dominated by large institutional investors, where transactions are based on bilateral negotiations and the main settlement method is real-time gross settlement. The exchange market, on the other hand, is more of a retail market dominated by smaller institutional investors, such as securities firms, funds, and high-net-worth individuals. The transactions on the exchange are often settled on a net basis.

Issuers

Credit bond issuance has been dominated by SOEs (Figure 4.7, panel 1). While the role of SOEs in China has declined significantly—in the past two decades, the share in employment fell below 15 percent and the share in industrial sales below 20 percent—SOEs still account for a disproportionally large share of credit, in both the banking system and the bond market (Lam and Schipke 2017). In 2017, SOEs accounted for 40 percent of outstanding bank loans, and 80 percent of outstanding credit bonds were issued by SOEs, particularly local SOEs, with variation across different bond types. For example, enterprise bonds are almost exclusively issued by SOEs, while the share of private firm issuance is close to 40 percent of corporate bonds and short commercial paper.

Figure 4.7.Issuer Ownership Structure, by Bond Type, 2017

(Percent)

Source: WIND Economic Database (www.wind.com.cn) 2017.

For enterprise bonds, LGFVs account for 80 percent of total issuance (Figure 4.7, panel 2). Given that these firms are often used to finance the off-budget spending of local governments, such as infrastructure investment, LGFV borrowings are classified as augmented government debt in the IMF definition (IMF 2018).6 Bonds issued by LGFVs often enjoy pricing advantages because they are generally considered to be backed by government guarantee. While the government has issued a number of regulations to break the implicit guarantees, the perception continues to be held by many market participants (see Chapter 11).

Investors

Credit bonds are almost entirely held by domestic investors. Reflecting various capital account restrictions, foreign participation in the credit bond market remains negligible, at 1 percent. Among domestic investors (Table 4.1), collective investment vehicles account for about 60 percent, followed by banks and securities firms. Reflecting the different trading platforms, securities firms have a higher participation rate in the corporate bond market (traded on exchanges) compared to other bonds that are mainly traded in the interbank market. Participation of long-term investors, such as pension and insurance funds, at 5 percent to 7 percent, is likewise limited.

Table 4.1.Credit Bond Domestic Investor Base, 2016(Percent)
Investor BaseCorporate BondsEnterprise BondsMedium-Term Notes
Collective investment vehicles646262
Banks182931
Securities firms1232
Insurance755
Clearing house010
Nonbank financial institutions010
Source: Authors’ compilation.
Source: Authors’ compilation.

Liquidity

Credit bond liquidity varies across different bond types, but is generally low. Liquidity is typically higher for shorter maturities, in particular commercial paper (partly reflecting maturities shorter than one year) and is also higher for bonds traded on the interbank market than those traded on the exchange market (Figure 4.8). For example, enterprise bonds, despite long maturities of 7–10 years, have higher liquidity than corporate bonds with shorter maturities. This is because the interbank market has attracted large institutional investors that contribute to more liquidity, while investors in the exchange market often use corporate bonds as a tool for liquidity management, with transactions dominated by collateralized repurchase agreements (repos).7 Overall, the liquidity of credit bonds in China is much lower than in advanced economies because most investors tend to hold paper until maturity.

Figure 4.8.Annual Turnover Ratio, 2017

Source: Bloomberg L.P. 2017.

Ratings

Bond ratings in China are highly skewed, reflecting both stringent issuance requirements and implicit guarantees. More than 95 percent of credit bonds are rated AA and above, compared to less than 6 percent in the United States (Figure 4.9). This partly reflects the stringent issuance requirements. For example, among corporate bonds, only bonds with AAA rating can be issued to all investors; for medium-term notes and commercial paper, only bonds with AA rating or above can be issued; for enterprise bonds, as noted, a compulsory external guarantee is required for firms with a leverage ratio above a certain threshold. This means that access to China’s credit bond market is quite restrictive and only firms with better quality can issue. At the same time, given the dominance of SOEs and the widespread market perception of government guarantees, the actual rating may overstate a firm’s underlying financial health. The implicit rating by China Bond, a central depository platform, shows that about 40 percent to 50 percent of credit bonds are rated AA and above, in contrast to the 80 percent in the official ratings (Figure 4.10).

Figure 4.9.Credit Bond Ratings Distribution: China and the United States, 2017

(Percent)

Source: Bloomberg L.P. 2017.

Figure 4.10.Official and Implicit Rating Distribution of Credit Bonds, 2017

(Percent)

Source: WIND Economic Database (www.wind.com.cn) 2017.

However, credit spreads in China are much higher for a given rating than in other countries. While the distribution of official ratings appears to be distorted, a closer look suggests that the same rating in China is associated with much higher credit spreads. For example, the credit spread of AAA bonds is typically 50 basis points in the United States and 87 basis points in China; for AA bonds, the credit spread is 175–220 basis points in China compared to only 70 basis points in the United States.8 Hence, a highly condensed rating distribution disguises the large variation of underlying credit spreads (Figure 4.11).

Figure 4.11.Credit Spreads, by Rating, China and the United States, 2017

(Basis points)

Source: WIND Economic Database (www.wind.com.cn) 2017.

Note: LGFV = local government financing vehicle.

Defaults

Credit bond defaults are a more recent phenomenon in China, and overall default rates remain low. Indeed, from the first issuance of credit bonds in 1983, almost no defaults occurred in China until 2014, when the economy slowed down visibly (Figure 4.12). The low default rate can be attributed to strong economic fundamentals and strict issuance approval, but also government protection of investors in near-default cases. From 2014, when the economy entered a structural slowdown, to 2016, the default rate rose from 0 percent to 0.5 percent—still significantly lower than the global average of 2 percent. In 2017, the rate fell to 0.2 percent as the economy entered a cyclical rebound (Figure 4.13).

Figure 4.12.Credit Bond Defaults, 2014–18

(Billions of renminbi, left scale; number of bonds, right scale)

Source: WIND Economic Database (www.wind.com.cn).

Figure 4.13.Credit Bond Default Rate in China and Globally, 2014–17

(Percent)

Sources: Brilliance Rating; and S&P Global.

Proper risk pricing of credit bonds is key to the efficient allocation of resources. As credit bonds become an increasingly important venue for the financing of firms, adequate risk pricing is critical to ensuring that credit resources are allocated efficiently, which affects the overall efficiency of the economy. Implicit guarantees, in particular of SOEs, may significantly distort risk pricing and be detrimental to the long-term development of the credit bond market. The next section addresses the extent to which credit bond prices reflect underlying risks.

Empirical Analysis

This section uses firm-level data to examine the pricing behavior of credit bonds in China. In particular, it analyzes the extent to which liquidity risk, credit risk, and the benchmark risk-free rate are reflected in bond prices and whether implicit guarantees have contributed to lower bond yields for SOEs and LGFVs.

In addition, given the segmentation of the bond market, it tests how pricing behavior differs across different credit bonds. Reflecting data restrictions, the analysis focuses on corporate and enterprise bonds that are traded on the exchange market. Medium-term notes and commercial paper are exclusively traded in the interbank market, and hence access to firm-level financial information is limited.

The literature on credit bond pricing in China is scant; the most recent study analyzes data up to 2012 (GAO and ZOU 2015) and hence does not capture the significant structural changes since then. In terms of methodology, considering the less-developed nature of the credit bond market in China, this chapter follows the linear approach used in GAO and ZOU (2015), and extends the analysis to bond market data until 2017.9 In comparison with Gao and Zhou (2015), one important deviation in the regression setup is that the analysis does not include credit ratings together with other firms’ fundamentals as explanatory variables, which may create collinearity issues. Another contribution to the literature is to quantify the impact of implicit guarantees on bond pricing.

The model is specified as follows, where the credit bond yield is regressed on the benchmark yield and a set of risk factors, while controlling for fixed effects.

The explanatory variables include the benchmark rate (Guokai yield) and risk factors (factor) (see annex 4.1 for detailed summary statistics on regression variables). In particular,

  • Benchmark rate: The analysis uses either the Treasury bond yield or the yield of the China Development Bank bond of the same maturity as the benchmark rate.10 The empirical results are robust to both specifications. While most studies use credit spread (bond yield minus benchmark) directly as the dependent variable, given the incomplete transmission in China, this chapter uses the benchmark rate as an explanatory variable to test the degree of transmission from the risk-free rate to the credit price.

  • Credit risk factors: A few variables are used to measure the credit risk of the issuer, including the size of total assets, interest coverage ratio, leverage ratio, and profit. The analysis also controls for ownership and firm type to test implicit guarantees for SOEs and LGFVs. Given that enterprise bonds are exclusively issued by SOEs, only data on corporate bonds (about 40 percent are issued by private firms) are used to test the pricing of implicit guarantees for SOEs.

  • Liquidity risk factors: The analysis uses the Amihud ratio,11 trading volume, and days to maturity to measure bond liquidity.

  • Systemic risk factors: Overall stock market return and year dummy are used to control for systemic macro risks.

Reflecting data availability, the analysis looks at enterprise and corporate bonds traded on the exchange market from 2011 to 2017, including monthly trading data for 2,319 enterprise bonds and 1,467 corporate bonds. It then matches them with the financial data of bond issuers, which are interpolated to monthly frequency from quarterly data, including leverage ratio, interest coverage ratio, asset size, and profit. A few clean-up steps were performed to drop missing values and extreme values, and to ensure minimum trading of each bond during the sample period. The final sample includes 318 enterprise bonds and 265 corporate bonds. (Annex 4.2 explains the data clean-up process and presents details about the calculations.)

Pricing in Different Markets

Transmission from the risk-free rate to credit bond yield has improved for enterprise bonds, but remains incomplete for corporate bonds. In the sample from 2011–14, while about 80 percent of adjustment in the risk-free rate was transmitted to both enterprise and corporate bond yields, in more recent years, the transmission mechanism has strengthened notably for enterprise bonds, with almost one-to-one pass-through. For corporate bonds, transmission remains partial, at close to 80 percent (Figure 4.14). This likely reflects the fact that most government and enterprise bond trading takes place in the interbank market, with significant overlap of the investor base contributing to better pass-through. Investors on the exchange market are typically smaller institutions with limited exposure to government bonds, and are hence less sensitive to benchmark rate changes. Lack of liquidity for corporate bonds has also contributed to the partial pass-through from the benchmark rate.

Figure 4.14.Response of Credit Bond Yield to Benchmark Rate, 2011–14 and 2015–17

(Transmission ratio)

Source: Authors’ estimates.

Corporate bonds have better pricing of credit risks than enterprise bonds, while neither are price sensitive to liquidity risks (Figure 4.15). The regression analysis shows that, for corporate bonds, a 1 percentage point increase in leverage pushes yields 3 basis points higher, and a 1 percentage point increase in profitability reduces bond yields by close to 2 basis points. For enterprise bonds, however, both leverage and profitability have negligible impacts on bond pricing. This likely reflects that most enterprise bond issuers are LGFVs and are closely linked to the government, implying that investors value the implicit guarantee more than the firm’s economic performance. In addition, external guarantees are often required for enterprise bond issuance for highly leveraged firms, which also contributes to the price insensitivity.

Figure 4.15.Bond Pricing of Credit and Liquidity Risks

(Percentage points)

Source: Authors’ estimates.

It is interesting that for both types of bonds, higher bond liquidity does not lead to lower yields. For corporate bonds, higher liquidity is even associated with higher spreads. This likely reflects the fact that most investors will hold the bond until maturity or use it for collateralized repos, and hence have less demand for trading and high liquidity.

Implicit Guarantees

The regression results show that the implicit guarantee of SOEs and LGFVs typically reduces bond yields by 100 basis points. After controlling for the type of credit bond, industry, and trading year, this analysis finds that SOEs pay 108 basis points less in yield than private firms with the same financial conditions (including leverage, profitability, and size). Similarly, the analysis finds that LGFVs enjoy a 97 basis point reduction in bond yields, reflecting the perceived guarantee from the government.

In the subsample, the analysis does not find significant reduction in the pricing of implicit guarantees. This shows that despite the implementation of the new budget law in 2015, investors still perceive LGFVs as closely connected to local governments and enjoying implicit guarantees. Such a perception has contributed to a crowding out of credit resources for private companies and lower credit efficiency.

Policy Recommendations

China’s credit bond market is still at an early stage of development and has significant room to grow. To foster a better allocation of resources and hence overall economic development, the bond market would benefit from the following:

  • Harmonize the regulation concerning different credit bond schemes. For historical reasons, China’s credit bond market is segmented, with different regulators, issuance procedures, trading platforms, and depositories. Harmonizing the regulation of the different bond schemes would be an important step forward to reduce segmentation and regulatory arbitrage, increase liquidity, and foster price discovery. It would also increase the attractiveness for foreign investors. Recently, the government established an interministerial committee led by the People’s Bank of China to work on unifying regulations. In September 2017, the regulations for rating agencies in the interbank market and exchange market were unified, which is a significant step and will pave the way for further harmonization in bond regulation.

  • Remove the implicit guarantee of SOEs and LGFVs. SOEs dominate China’s overall credit bond issuance, and on average pay 100 basis points less in bond yields than private firms with similar financial conditions, reflecting the implicit guarantees associated with SOEs. Investors still perceive LGFVs as having implicit guarantees, despite regulatory tightening. Such pricing distortions crowd out dynamic private issuers and lead to inefficient allocation of credit resources, while creating contingent government liabilities down the road. To break the implicit guarantee, the government should allow the exit of weak SOEs and LGFVs, instead of stepping in with last-minute government support. In the meantime, such removal must proceed at a managed pace because a sudden break of implicit guarantees may lead to dramatic asset repricing and financial stability risks.

  • Broaden the investor base. The credit bond market is dominated by domestic investors and needs to open up to attract foreign institutional investors. A more diversified investor base could promote better risk pricing and contribute to higher liquidity in the bond market. Other measures with respect to market structure, including promoting the outright repo market and centralizing market making, will also help increase bond liquidity (see Chapter 12).

  • Align the rating system with global standards. Current credit bond ratings appear to be distorted compared to international standards. This certainly reflects the stringent regulatory threshold, but also reflects a nascent rating industry, which needs further improvement to provide more informative ratings. Bringing in more foreign participation in the ratings industry could promote a better risk culture and improve the quality of bond ratings. The licensing of three international rating agencies in 2018 is an important step forward.

  • Promote better risk sharing between investors and bond issuers. Standards to issue credit bonds in China are stringent and issuance often requires approval. This partly reflects the regulator’s intention to protect investors from potential default. However, more than 99 percent of bond investors are professional institutional investors who should be able to bear a certain degree of default risk. Regulators may strike a better balance in risk sharing between investors and bond issuers by lowering the regulatory threshold for bond issuance and encouraging more issuance by smaller private firms, while allowing bonds to default and investors to bear the risks.

Conclusions

China’s credit bond market has developed rapidly in the past decade and is now the second largest in the world. Nonetheless, bond financing still accounts for only 10 percent of nonfinancial corporate financing, compared to around 60 percent in the United States. In addition, foreign investors currently hold less than 1 percent of credit bonds. This implies significant room for further bond market development and increased foreign participation.

The current fragmented credit bond market structure is the result of the market’s unique historical development. New bond schemes (such as corporate bonds and medium-term notes) were introduced and developed in parallel to old schemes (such as enterprise bonds), which to some extent reflects the difficulty of fully reforming the old scheme for historical reasons. Different segments of credit bonds are also associated with different regulators and different trading and settlement platforms. The market is dominated by SOEs and LGFVs, and issuance is highly restricted. In terms of ratings, 90 percent of credit bonds are rated above AA.

Based on firm-level data, the empirical analysis in this chapter provides unique insights into China’s credit bond markets. Credit risk, for example, is more adequately reflected in the pricing of corporate bonds compared to similar enterprise bonds. At the same time, there is little pricing sensitivity to liquidity risks. The analysis also reveals that SOEs and LGFVs benefit from lower financing costs, amounting to a reduction of some 100 basis points when compared to private firms with similar financial and operating conditions. This reduction reflects the existence of implicit guarantees.

To take advantage of the full potential of China’s credit bond market and improve resource allocations, further reforms will be needed:

  • First, there is significant room to unify issuance procedures and regulations across different credit bond schemes. Over a longer horizon, a discussion about the benefits and costs of unifying the different regulators might be warranted.

  • Second, to foster a more efficient allocation of resources, it will be important to eliminate government guarantees. Measures put in place in 2017–18 to break the link between local governments and LGFVs are already an important step in this direction. However, to create a level playing field, it will also be important to allow nonperforming SOEs to exit the market and hence allow corporate bond defaults (see Chapter 11).

  • Third, efforts are needed to broaden the investor base and to attract more institutional investors into the market, including further opening up to foreign investors.

  • Fourth, the rating industry in China needs further development to align with global standards. The approval of foreign rating companies in 2018 to enter the market may play an important role in this regard. Following a strengthened credit culture, the government could consider lowering the entry threshold for bond issuance and give small and medium private firms more access to bond issuance, with higher yields that properly reflect their underlying risks.

With these reforms, the credit bond market could become a more important venue for corporate financing and promote economic growth by allocating credit resources efficiently.

Annex 4.1. Regression Variables
Annex Table 4.1.1.Regression Variables
CharacteristicsTypeVariablesUnitExplanationExpected Sign
Bond yieldDependent variableBond yield%Monthly bond yields
Baseline interest rateIndependent variableRisk-free rate%China Development Bank bond yield with same maturityClose to 1
Credit risk measurementIndependent variableSOEDummySOE = 1,non-SOE = 0-
Independent variableLGFVDummyLGFV=1,non-LGFV = 0-
Independent variableTotal assetLog RMB millionNatural log of the total assets of the issuance company-
Independent variableProfit%Net profit/revenue-
Independent variableInterest coverage ratio%Interest coverage ratio-
Independent variableLeverage ratio%Total Liability/Total Asset x 100%+
Liquidity risk measurementIndependent variablellliquidity measure%/RMB millionAmihud (2002), sum of monthly return/ monthly trade dollar volume+
Independent variableTrading volumeLog RMB millionAverage daily trading volume-
Independent variableDays left to maturityDayMaturity day-trading day-
Bond specificsIndependent variableIndustryDummyDifferent industries
Independent variableBond issuance lengthYearsBond issuance length
Independent variableIssuance sizeRMB 100 millionBond issuance size
Systematic risk measurementIndependent variableStock market change%China Securities Index 300 index change-
Independent variableTrading yearDummyYear of the trading day
Source: Authors' compilations.Note: CSI = China Securities Index; ICR = interest coverage ratio; LGFV = local government financing vehicle; RMB = renminbi; SOE = state-owned enterprise.
Source: Authors' compilations.Note: CSI = China Securities Index; ICR = interest coverage ratio; LGFV = local government financing vehicle; RMB = renminbi; SOE = state-owned enterprise.
Annex 4.2. Data Cleaning and Summary Statistics

Annex Figure 4.2.1.Data Cleaning

Source: Authors' compilations.

Note: RMB = renminbi.

Annex Table 4.2.1.Summary Statistics: Enterprise Bonds
VariableNumber of ObservationsMeanStandard DeviationMinimumMaximum
Bond yield16,0395.21.5-13.714.6
Risk-free rate28,1074.20.72.66.4
LGFVs28,4540.80.40.01.0
Total assets27,11624.61.619.629.0
Profits26,24321.521.6-7.178.7
Interest coverage ratio23,56416.131.60.8130.6
Leverage ratio27,11652.016.60.395.2
Illiquidity measure8,450-0.48.2-26.721.7
Trading volume28,1071.93.10.09.0
Days left to maturity28,1072,6081,0983878,218
Bond issuance length28,4547.92.7330
Issuance size28,45420.218.810200
Stock market change28,1070.57.1-21.025.8
Trading year28,10720141.9520112017
Source: Authors’ calculations.Note: LGFV = local government financing vehicle.
Source: Authors’ calculations.Note: LGFV = local government financing vehicle.
Annex Table 4.2.2.Summary Statistics: Corporate Bonds
VariableNumber of ObservationsMeanStandard DeviationMinimumMaximum
Bond yield22,1941.52.40.014.7
Risk-free rate22,1944.20.72.66.3
SOE dummy22,4680.50.50.01.0
Total assets20,47624.91.416.429.1
Profits20,32312.310.40.639.2
Interest coverage ratio19,0839.312.61.251.8
Leverage ratio5,60564.115.00.6101.3
Illiquidity measure4,7130.00.0-0.10.1
Trading volume22,1941.53.00.09.6
Days left to maturity22,4682,0296621,0955,478
Bond issuance length22,4685.61.8315
Issuance size22,46826.520.010.0160.0
Stock market change22,1940.57.1-21.025.8
Trading year22,19420141.9520112017
Source: Authors’ calculations.Note: SOE = state-owned enterprise.
Source: Authors’ calculations.Note: SOE = state-owned enterprise.
Annex 4.3. Regression Results

The following table presents the sensitivity of credit bond yields to the benchmark rate, credit risks, liquidity risks, and ownership characteristics. The pricing behavior is estimated separately for corporate bonds and enterprise bonds using both the ordinary least squares and fixed-effects methods.

Annex Table 4.3.1.Sensitivity of Credit Bond Yields
Enterprise BondsCorporate Bonds
Model (1a)Model (1b)Model (2a)Model (2b)
VariablesOLSFixed EffectsOLSFixed Effects
Risk-free rate (CDB bond yield)0.797***

-0.0334
0.769***

-0.0333
0.714***

-0.0645
0.736***

-0.0612
LGFV dummy-0.968***

-0.233
SOE dummy-1.081***

-0.13
Total assets-0.304***

-0.0997
-0.164

-0.241
-0.746***

-0.149
-1.302***

-0.31
ICR0.00225

-0.00202
0.00243

-0.00216
0.00678**

-0.00274
0.00814***

-0.0028
Leverage ratio (Debt/Assets)-0.00235

-0.00463
-0.00423

-0.00671
0.0321***

-0.01
0.0389***

-0.0128
Illiquidity measure (Amihud ratio)-0.00951***

-0.00149
-0.00967***

-0.0015
-2.387***

-0.417
-2.412***

-0.409
Profit (net profit margin)-0.00151

-0.00162
-0.00156

-0.00172
-0.0145***

-0.00411
-0.0152***

-0.00454
Trading volume0.0129

-0.013
0.00584

-0.0129
0.0640***

-0.0113
0.0624***

-0.0111
Days left to maturity (bond life)0.000166**

-8.22E-05
0.000791***

-0.000142
-9.76E-05

-0.000123
-0.00027

-0.000197
Issuance size-0.00587

-0.00486
0.00668

-0.00595
Bond maturity-0.0821***

-0.0208
0.0527

-0.0467
Stock market exchange0.00263***

-0.00097
0.00389***

-0.00098
-0.00141

-0.00111
-0.00092

-0.00112
Constant10.87***

-2.451
4.342

-6.028
19.18***

-3.259
32.78***

-7.595
Control for yearYesYesYesYes
Control for industryYesYes
No. of observations7,4027,4024,3134,313
R20.5590.493
No. of bonds318318265265
Source: Authors’ calculations.Note: CDB = China Development Bank; ICR = interest coverage ratio; LGFV = local government financing vehicle; OLS = ordinary least squares; SOE = state-owned enterprise.***p < .01; **p < .05.
Source: Authors’ calculations.Note: CDB = China Development Bank; ICR = interest coverage ratio; LGFV = local government financing vehicle; OLS = ordinary least squares; SOE = state-owned enterprise.***p < .01; **p < .05.
References

    AmihudY. 2002. “Illiquidity and Stock Returns: Cross-Section and Time-Series Effects.” Journal of Financial Markets5 (1): 3156.

    DufeeG. R.1999. “Estimating the Price of Default Risk.” Review of Financial Studies12 (1): 197226.

    EomY. H.J.Helwege andJ. Z.HUANG. 2004. “Structural Models of Corporate Bond Pricing: An Empirical Analysis.” Social Science Electronic Publishing17 (2): 499544.

    GAOQ. andH. F.ZOU. 2015. “A Comparative Study on the Secondary Market Pricing of China’s Enterprise Bonds and Corporate Bonds.” Journal of Financial Research1: 84100 (in Chinese).

    GündüzY. andM.Uhrig-Homburg. 2014. “Does Modeling Framework Matter? A Comparative Study of Structural and Reduced-Form Models.” Review of Derivatives Research17 (1): 3978.

    International Monetary Fund (IMF). 2018. “China Article IV Staff Report August 2018.” IMFWashington, DC.

    JarrowR. A. andS. M.Turnbull. 1995. “Pricing Derivatives on Financial Securities Subject to Credit Risk.” Journal of Finance50 (1): 5385.

    LamW.RaphaelM.Rodlauer andA.Schipke. 2017. Modernizing China: Investing in Soft Infrastructure. Washington, DC: International Monetary Fund.

    LiK. L. andH. Y.Wong. 2008. “Structural Models of Corporate Bond Pricing with Maximum Likelihood Estimation.” Journal of Empirical Finance15 (4): 75177.

    MertonR. C.1974. “On the Pricing of Corporate Debt: The Risk Structure of Interest Rates.” Journal of Finance29 (2): 44970.

Bond Connect is a new mutual market access scheme that allows investors from mainland China and overseas to trade in each other’s bond markets through connection between the related mainland and Hong Kong SAR financial infrastructure institutions.

This chapter focuses only on credit bonds issued by nonfinancial firms.

For example, for enterprise bond issuance, a compulsory explicit guarantee is required for issuers with debt-to-assets ratios of 75 percent or above, except for AAA issuers with ratios below 85 percent and AA+ issuers with ratios below 80 percent. The outstanding medium-term notes cannot exceed 40 percent of a firm’s net assets.

In addition to commercial paper and medium-term notes, nonfinancial-firm financing instruments also include asset-backed notes and private placement notes.

In addition, less than 0.3 percent of transactions take place in the commercial bank over-the-counter market and the free trade zone bond market.

In the 2014 national audit, the Chinese government officially acknowledged about 13 trillion renminbi in LGFV debt as government debt and since then has implemented various measures to break the link between LGFVs and government debt.

In this case, corporate bonds are used as collateral for repo transactions.

The high-yield AA bond in China, however, still has a lower spread than the high-yield bond in the United States, which has not dropped below 300 basis points since 2007.

The literature is rich in structural models for credit pricing, including the Merton model, the Black and Cox model, the Collin-Dufresne model, and the Longstaff-Schwartz model. However, these models are not ideal for empirical testing of credit pricing (Merton 1974; Eom, Helwege, and HUANG 2004; Li and Wong 2008). The reduced form model (Jarrow and Turnbull 1995) is an alternative method, but the estimation is not robust (Dufee 1999) and often has a high estimation error similar to structural models (Gündüz and Uhrig-Homburg 2014). The linear model can easily include credit, liquidity, and systematic risk in bond pricing analysis and has been widely used empirically in recent studies (GAO and ZOU 2015).

The China Development Bank bond yield is the de facto benchmark rate in China’s credit bond market, given its sovereign-grade ratings and much higher liquidity than the sovereign bond.

The Amihud ratio is calculated as monthly return divided by monthly trading volume. The higher the ratio, the lower the liquidity.

    Other Resources Citing This Publication