Issuance overpricing of China's corporate debt securities

Abstract

We document issuance overpricing of corporate debt securities in China, which is robust across subsamples with different credit ratings, maturities, and issuers. This phenomenon contrasts with underpricing of equity and debt securities in Western countries and reflects China's distinct institutional environment. The average overpricing dropped from 7.44 basis points to 2.41 basis points after the government prohibited underwriters from using rebates in issuances in October 2017. By analyzing overpricing before and after the rebate ban and across different issuers and underwriters, we uncover two channels for underwriters, who compete for future underwriting business, to drive up overpricing: rebates and self-purchases.

Introduction

Due to the Chinese government's efforts to liberalize financial markets, China's corporate debt-security market has grown rapidly in recent years. According to SIFMA (https://www.sifma.org) and the Asian Development Bank (https://asianbondsonline.adb.org), by the end of 2019, outstanding corporate debt securities in China reached 4.3 trillion USD, making it the second-largest market of corporate debt securities in the world, just behind the U.S., with 10.6 trillion USD.¹ China's corporate debt-security market is different from the more developed U.S. market in several ways, as reviewed by Amstad and He (2020). First, it directly grew out of China's banking sector and features banks as its major investors and underwriters. Consequently, issuance of debt securities directly competes with bank loans for firms’ financing needs. Second, debt securities issued in this market are mostly commercial paper and medium-term notes with an average maturity of 1.74 years, which is substantially shorter than maturities in the U.S. market. Third, debt securities in China tend to be issued by large firms with credit ratings highly skewed to the upside and artificially low default rates, which possibly reflects the government's tight control on issuances, as well as implicit government guarantees to avoid public defaults.² These differences make studying how pricing and market dynamics in China's newly developed and rapidly growing market of corporate debt securities may be different from that in other countries particularly interesting and important.

In this paper, we focus on the issuance pricing of debt securities in China's interbank market, which accounts for about 90% of debt securities issued in China in recent years.³ We collect a comprehensive data set of 18,229 debt securities issued by 2,558 nonfinancial firms in 2015–2019 in the interbank market, including both initial and seasoned offerings. We uncover strong evidence of issuance overpricing: the yield spread of a debt security on the day of its first secondary-market trade is, on average, 4.9 basis points (bps) higher than its yield spread at the issuance, relative to the yield of a Treasury security with similar maturity.⁴ This overpricing is robust across debt securities with different characteristics, such as initial or seasoned offering, maturity, and underwriter type, and across issuers with different attributes, such as credit rating, size, and state ownership.

This issuance overpricing is in sharp contrast to the typical observations of issuance underpricing of both equity and debt securities in the U.S. and other countries. See Lowry et al. (2017) for a review of the extensive evidence of underpricing in equity initial public offerings (IPOs). Although the literature on the issuance pricing of corporate debt securities, which we review later, is less conclusive, it mostly finds evidence of underpricing of corporate debt securities in developed economies.

The pervasive issuance overpricing reflects the different institutional environment in China and thus offers a window to examine the second-largest market for corporate debt securities in the world. In this market, banks with ample investment capital act as both underwriters and investors to compete for the issuances of large firms with relatively low default risks. Interestingly, they compete on issuance pricing. Note the secondary market for debt securities tends to be highly illiquid, which makes the secondary-market price more manipulable and thus less reliable than the issuance price. The illiquidity of the secondary market also makes new issues more appealing to investors than purchases in the secondary market despite the issuance overpricing. By contrast, the issuance price is not only more reliable than the secondary-market price, despite the issuance overpricing, but also regularly available to the public. Due to the short maturities in China's debt-security market, most firms need to repeatedly issue debt securities. As a result, higher pricing not only reduces the issuer's financing cost of the current issuance, but also provides a publicly observed benchmark for the issuer's other debt financing, such as bank loans. The benchmark role of the issuance price thus induces an issuer to reward its future issuance to its current underwriter based on issuance pricing, rather than on price stability in the secondary market or on the underwriting fee. Consistent with this incentive, we find evidence in the data that a lower yield spread in the issuance (i.e., higher pricing) predicts a higher probability of the issuer's retention of its current underwriter for its subsequent issuance.

How do underwriters generate overpricing? The interbank market in China offers different channels for the underwriter to influence the pricing. In the U.S., a syndicate allocates the issuance of a corporate bond among potential investors and usually sets the offering price below the level that is expected to prevail in secondary-market trading to induce investors to reveal their demand. Furthermore, the syndicate assumes the obligation to stabilize the issuance in the secondary market should demand prove to be weaker than expected, as discussed in Bessembinder et al. (2020). In contrast, the issuance of corporate debt securities in China entails a single-price auction in which the underwriter and other qualified institutions directly bid the issuance for themselves or their clients. The underwriter does not allocate the issuance and instead serves to organize the auction and contact potential investors to participate in it. Furthermore, the underwriter is not obligated to support the issuance in the secondary market.

A simple channel through which the underwriter can affect the pricing is to offer rebates to attract participants to the auction. Because underwriters do not need to disclose the rebates to the public, they can use rebates to price discriminate investors, potentially corrupting the transparency and quality of the issuance process. The regulator of China's interbank market, the National Association of Financial Market Institutional Investors (NAFMII), was so concerned about the widespread use of rebates that it issued a new regulation banning underwriters from using rebates after October 1, 2017. The average issuance overpricing dropped from 7.44 bps before the rebate ban to 2.41 bps after the rebate ban. This policy shock allows us to further examine the effects of rebates and the underwriter's incentives for issuance overpricing.

Because a stronger incentive for the underwriter to win the issuer's future business induced the underwriter to use rebates to generate a higher overpricing before the rebate ban, we hypothesize that issuances by underwriters with stronger incentives experienced greater drops in overpricing after the rebate ban. This hypothesis motivates a difference-in-differences analysis of how the drop in overpricing varies across different issuers and underwriters. Because central state-owned enterprises (SOEs) are usually giant firms that enjoy the central government's implicit guarantees, they are more valuable issuers than other firms and thus attract more intense competition for their issuances. Consistent with this notion, we find that after the rebate ban, the drop in overpricing was significantly greater for debt securities issued by central SOEs than for those by other issuers. Furthermore, because the “Big Four” banks (Industrial and Commercial Bank of China, Agricultural Bank of China, Bank of China, and China Construction Bank) are the largest underwriters in the interbank market, these top underwriters face less competition and thus have fewer incentives to use rebates to generate overpricing. We indeed find that the drop in overpricing after the rebate ban was significantly smaller for issuances underwritten by the Big Four banks. These significant difference-in-differences results support our hypothesis and highlight underwriters’ incentives and their use of rebates as an important mechanism for generating issuance overpricing.

Interestingly, even after the rebate ban, issuance overpricing remained significant, albeit reduced, suggesting additional forces at work to sustain overpricing. As a key feature of China's interbank market, most licensed underwriters are banks, which regularly purchase corporate debt securities in the primary market for their own investment accounts.⁵ Indeed, our data show that underwriters, on average, acquire 35% of the debt securities that they underwrite themselves either for their own accounts or clients. We find a surprising pattern that underwriters tend to acquire more in issuances with higher overpricing, suggesting underwriters take losses in their self-purchases. This pattern contradicts two alternative hypotheses. One posits that underwriters acquire the issued securities to take advantage of market undervaluation and thus earn superior returns from their self-purchases. The other argues that underwriters provide price support to the issuances at the fundamental values, which implies fair returns for their self-purchases. Instead, this finding points to overbidding by underwriters as another channel for generating issuance overpricing, which is particularly relevant after the rebate ban.⁶

Finally, we confirm that the rebate ban has helped improve the quality of the issuance price. By examining the ability of a set of publicly observed debt and issuer characteristics to explain the cross-sectional variation of the issuance price before and after the rebate ban, we find that the explanatory power of these fundamental variables increased after the rebate ban, indicating an improvement in the quality of the issuance price.

Overall, our study not only documents pervasive issuance overpricing in China's interbank market for corporate debt securities, but also attributes this surprising finding to China's distinct institutional environment and issuance process. Our study thus motivates future studies to systematically examine the implications of the distinct institutional arrangements in this market on other important aspects, such as asset pricing and market efficiency, and to compare China with Western countries on these aspects.

The related literature

Our study contributes to several strands of the finance literature. First, it expands the literature on the issuance pricing of corporate debt securities. Datta et al. (1997), Helwege and Kleiman (1998), Cai et al. (2007), Hale and Santos (2007), Goldstein and Hotchkiss (2009), and Bessembinder et al. (2020) show significant underpricing for IPOs and seasonal offerings of corporate bonds in the U.S., although their findings on investment-grade bonds are mixed.⁷ Evidence of underpricing in the European markets also exists, for example, Wasserfallen and Wydler (1988) and Zaremba (2014). Different from other major markets of corporate debt securities around the world, Matsui (2006) and McKenzie and Takaoka (2008) provide preliminary evidence of issuance overpricing in Japan. Perhaps unsurprisingly, overpricing appears both in Japan and China because their markets share important similarities, including having banks as major investors and underwriters. Our analysis not only shows robust evidence of issuance overpricing in China, but also provides extensive analysis of the mechanisms that lead to the overpricing, which is missing from the studies of overpricing in Japan.

Second, our paper adds to the economic understanding of issuance pricing. The existing literature focuses mostly on two key mechanisms for security underpricing. One is information asymmetry, for example, Rock (1986), Benveniste and Spindt (1989), Benveniste et al. (2002), and Sherman and Titman (2002), and the other is liquidity, for example, Booth and Chua (1996) and Ellul and Pagano (2006). Complementing these studies,⁸ our analysis provides extensive evidence for a different mechanism that generates issuance overpricing, specifically, underwriter competition. Datta et al. (1997) speculate that the overpricing they find in their study of bond IPOs in the U.S. market could be driven by excessive competition among underwriters, but they do not provide evidence on either underwriter incentives or the channels.

Third, our paper also complements the literature on how business relationships affect the book-building process. Several studies document that in the U.S. market, underwriters may misuse their discretion over both price and the allocation of new issuances. Instead of underpricing the issuances to reward information production, underwriters may use underpricing as a quid pro quo for the investors’ future businesses or kickbacks, with new issuances being allocated based on a continuing business relationship (e.g., Reuter, 2006; Nimalendran et al., 2007; Liu and Ritter, 2010; Goldstein et al., 2011). Our study shows that under China's institutional environment, future business relationships between the underwriter and the issuer lead to issuance overpricing rather than underpricing.

Finally, our paper also adds to the quickly growing literature on China's financial system. See the handbook edited by Amstad et al. (2020) for chapters on different segments of China's financial system and, in particular, the chapter by Amstad and He (2020) for an overview of China's interbank market for corporate debt securities. Ang et al. (2017) examine the pricing of municipal bonds in China and link the pricing to real estate and political risks. Chen et al. (2020) argue that the rapid growth of China's municipal bond market is driven by the need of local governments’ financing platforms to roll over bank loans initially given during China's 4 trillion RMB post-crisis stimulus package. Chang and Liu (2021) propose a theory to explain the regulator’ incentive for developing bond market and, more broadly, allowing the growth of shadow-banking system, which is to contain the overinvestment caused by local government. By exploring the different rules used by the interbank market and the exchange market for repo transactions, Chen et al. (2019) find that an increase in the haircut requirement can have a substantial effect on firms’ debt-financing costs in China. Liu et al. (2017), Geng and Pan (2020), and Chang et al. (2021) discuss the impact of implicit guarantee on bond pricing. Our paper shares the common theme of these papers in exploring important characteristics of China's corporate debt-security market, but with a distinct focus on issuance pricing.

The remainder of the article is organized as follows. Section 2 introduces the institutional background of China's interbank market. Section 3 summarizes our data and measurement methodology. Section 4 documents issuance overpricing, and Section 5 examines the economic mechanisms. We conclude the paper in Section 6. We also provide an Internet Appendix, which is available at the journal website, to report additional results.