تاثیر تجارت CDS بر بازار اوراق قرضه: شواهدی از آسیا

This paper investigates the impact of CDS trading on the development of the bond market in Asia. In general, CDS trading has lowered the cost of issuing bonds and enhanced the liquidity in the bond market. The positive impact is stronger for smaller firms, non-financial firms and those firms with higher liquidity in the CDS market. These empirical findings support the diversification and information hypotheses in the literature. Nevertheless, CDS trading has also introduced a new source of risk. There is strong evidence that, at the peak of the recent global financial crisis, those firms included in CDS indices faced higher bond yield spreads than those not included.

The rise and fall of the credit derivatives market have been considered as the single most important development in the global credit market in the past decade. Before the onset of the global financial turmoil that started in mid-2007, the use of credit default swaps (CDSs) as an instrument to trade credit risk had increased exponentially. Since 2008, however, activity in the CDS market has shrunk substantially. In particular, Fig. 1 shows that the CDS notional amount outstanding dropped from roughly $60 trillion at the end of 2007 to about $33 trillion at the end of 2009, reflecting severely strained credit markets and the increased multilateral netting of offsetting positions by market participants.1 Full-size image (16 K) Fig. 1. Global CDS markets: Notional amounts outstanding. Source: Bank for International Settlements. Figure options In Asia, the CDS market has also grown rapidly in the past decade, despite the fact that it is relatively small and illiquid compared to its counterparts in Europe and the United States. Table 1 shows the gross notional amount outstanding and the number of contracts of major CDS indices covering reference entities in North America, Europe and Asia as of April 2011. CDS contracts written on Asian reference entities started to be traded in the late 1990s (Remolona and Shim, 2008). In July 2004, the first CDS indices focusing on the region, iTraxx Japan and iTraxx Asia ex-Japan, started to be traded in the market. The trading was relatively limited in the first few years, and then picked up strongly after these indices were reconstituted to reflect a surge in bond issuance by new large borrowers in the region starting in the fourth quarter of 2006. The gained liquidity in the index market spilled over into single-name CDS contracts. As a result, the CDS market in Asia began to emerge as a potentially serious market in its own right. Table 1. CDS index market by region. Dealer Non-dealer/customer Total Notional Contracts Notional Contracts Notional Contracts North America 2797 32,429 1084 10,268 3881 42,697 Europe 3834 39,820 1070 11,972 4904 51,792 Asia 169 9642 18 1213 187 10,855 “Notional” is the gross notional amount outstanding of major CDS indices in billion US dollars as of the week ending on 1 April, 2011, and “Contracts” is the number of CDS index contracts outstanding as of the same period. CDS indices for North America include CDX.NA.HY.15 index (on-the-run), CDX.NA.IG.16 index (on-the-run), CDX.NA.HY index (off-the-run) and CDX.NA.IG index (off-the-run). CDS indices for Europe include ITRAXX EUROPE CROSSOVER SERIES 15 (on-the-run), ITRAXX EUROPE SERIES 15 (on-the-run), ITRAXX EUROPE (off-the-run) and ITRAXX EUROPE CROSSOVER (off-the-run). CDS indices for Asia include ITRAXX ASIA EX-JAPAN IG SERIES 15 (on-the-run), ITRAXX JAPAN SERIES 14 (on-the-run), ITRAXX ASIA EX-JAPAN (off-the-run), ITRAXX ASIA EX-JAPAN HY (off-the-run), ITRAXX ASIA EX-JAPAN IG (off-the-run) and ITRAXX JAPAN (off-the-run). The categorisation of “Dealer” and “Non-dealer/customer” is by type of sellers of protection. Source: DTCC. Table options Compared to Europe and the United States, the CDS market in Asia was introduced against the backdrop of a corporate bond market still in its infancy. Fig. 2 shows the amount outstanding of debt securities issued by financial and non-financial corporate issuers residing in the United States, Europe and Asia from 1993 to 2011. In most Asian economies, the size of the bond market was quite small, the trading was not liquid, and the issuance of new bonds was largely driven by quasi-government issuers or issuers with some form of credit guarantees. Since the mid-2000s, however, the depth and breadth of the Asian bond market have improved greatly, due partly to the priority focus of Asian policymakers to develop local currency bond markets and partly to the positive spillover between the bond market and the derivatives market (Gyntelberg et al., 2005 and Filardo et al., 2010). Full-size image (26 K) Fig. 2. Corporate debt securities outstanding by residence of issuers. The outstanding amount of domestic and international bonds and notes issued by financial firms and non-financial firms located in each region. The data is quarterly, starting from Q4 1993 and ending in Q3 2011. Source: Bank for International Settlements. Figure options For policymakers, an issue of interest is the impact of the development of the CDS market on the corporate bond market. There are two approaches in the literature. One is to examine both the bond and CDS markets and draw comparisons about their roles in price discovery. The general results suggest that CDSs play a leading role in price discovery (Baba and Inada, 2009, BIS, 2003, Blanco et al., 2005, Hull et al., 2004 and Zhu, 2006), although there is a no-arbitrage relationship between the two markets in the long run.2 The other approach focuses on the bond market and examines whether bond market development, e.g., the cost of issuing and trading bonds, is affected by the introduction of CDS markets (Ashcraft and Santos, 2009). In this paper, we adopt the second approach and leave the first issue untouched. The limited number of studies on the impact of CDS trading on the bond market have so far focused on the US market. However, their findings do not necessarily apply to the Asian market. One reason for this is the different stages of development between the Asian and US bond markets when CDSs were introduced. Given that the bond markets in many Asian economies were underdeveloped from the beginning, it is more likely that CDS trading will have a jump-start effect on bond market development. Moreover, investors in Asian bond markets, especially in local currency bond markets, are predominantly domestic investors, while those in Asian CDS markets are mostly foreign investors. Baba and Inada (2009) document that the main investors in the Japanese CDS market are non-Japanese hedge funds and investment banks, while those in the Japanese corporate bond market are Japanese institutional investors such as life insurance companies and pension funds. The different pool of investors is likely to be an important channel to enhance information transparency and improve efficiency in the credit market. This paper attempts to investigate the empirical linkages between the bond and CDS markets in Asia during the period from January 2003 to June 2009. In particular, we address the following three questions. First, what is the impact of CDS trading on the Asian bond market in terms of issuance cost and liquidity? Second, which subset of bond issuers is most likely to benefit from the trading in the CDS markets? Third, did the impact of CDS trading on the bond market exhibit new characteristics during the recent global financial crisis? The main findings are as follows. First, we find strong evidence that CDS trading is associated with lower cost and higher liquidity for new bond issuance in Asia. This is consistent with the hypothesis that CDS trading helps create new hedging opportunities and improve information transparency for investors. Noticeably, this result is contrary to similar studies based on US data. This contrast provides supporting evidence for our conjecture on the jump-start effect in Asia. Second, we find that the positive impact of CDS trading on the bond market tends to be more remarkable for smaller firms and non-financial firms. In addition, those firms with higher liquidity in the CDS market benefit more in the primary bond market in terms of cost and liquidity. Last, we also find that the impact of CDS trading on the bond market is different during the crisis period. The global financial crisis that occurred during the sample period offers a good case study to examine the behaviour of the CDS and bond markets under distress and their linkages. Our analysis shows that, at the peak of the global financial crisis, those firms included in CDS indices had to face higher spreads than those not included in CDS indices, above and beyond the general increase in credit spreads observed in the bond market during this period. This suggests that CDS trading could be a double-edged sword: it also introduces new sources of shocks to the bond market. The rest of the paper is organised as follows. Section 2 reviews the literature and highlights the contributions of this study. Section 3 documents the empirical methodology. Section 4 describes data, and Section 5 reports empirical results related to the above three questions. Finally, Section 6 concludes with discussion of policy implications.

This paper shows that in Asia, CDS trading has had positive impacts on bond market development in terms of lowering average spreads and enhancing the market liquidity before and in the early stage of the recent international financial crisis. This finding supports the diversification and information hypotheses, and justifies continued development of active CDS markets in the region. For financial regulators, the sequencing problem, i.e., the order in which various financial markets should be developed, is an important issue (Chami et al., 2010). Historically, derivatives markets have been introduced well after the development of the market for the underlying securities. Once derivatives become available, dealers and bond holders can hedge their exposures much more easily and efficiently. In addition, derivatives facilitate the pricing of instruments by making it easier to repackage risks and exploit arbitrage opportunities. This makes the holding of bonds more attractive and contributes to the development of secondary markets in the underlying bonds. Also, as the experience of derivatives on US government bonds shows, the existence of derivatives can significantly increase the liquidity of the market for the underlying bonds and encourage development of the bond market. Regulators might consider promoting development of the CDS market, so that the CDS market continues to lower the costs and increase liquidity in the bond market. However, we also find that at the peak of the financial crisis, the CDS market contributed to higher spreads in the bond market. Given that the CDS market plays the role of shock amplifier during the crisis, it is important to make sure that policymakers introduce measures to mitigate the negative spillover effect from the CDS market to the bond market. Rapid growth and lax supervision of the CDS market up to 2007 have raised a number of policy concerns. The current regulatory reform initiatives on the CDS market aim at reducing systemic risk and providing additional transparency in the credit derivatives market. Most discussions have focused on how to reduce systemic risk stemming from the widespread use of CDS contracts traded in the over-the-counter (OTC) market. The Financial Stability Board and the Group of Twenty have picked up this issue, and agreed to require that all standardised CDS contracts trade in principle through central counterparties (CCPs) or exchanges. The US Treasury Department decided to require all standard credit default swaps to be traded through a CCP, and a similar undertaking is being contemplated in Europe. The current discussion seems to focus on financial stability benefits. Trading through CCPs has certain advantages over trading in OTC markets. Counterparty risk and settlement risk are important considerations in OTC markets, whereas CCPs can limit delivery failures by requiring accounts to be marked-to-market and by enforcing margin or collateral requirements. Nevertheless, as Stulz (2010) suggests, the use of clearing houses is not a panacea for eliminating systemic risk associated with OTC trading of derivatives. Also, clearing houses are inefficient at dealing with most new financial products and customised derivatives. The OTC market is better at enabling innovation, at addressing specific derivatives requirements from end-users, and at finding counterparties when liquidity for a derivative on an exchange is low. By contrast, exchanges are more efficient when there is a large volume of trading for standardised contracts. In this sense, as emphasised by Duffie (2009), it is important to improve price transparency and efficiency in the credit derivatives market. Trading through CCPs could be helpful in achieving this objective. However, given that the credit market, including the CDS market, is relatively less developed in Asia, the optimal strategy for setting up CCPs remains an open issue. To this end, a continued discussion is warranted to strike the right balance between maintaining the stability of the CDS markets and promoting efficient and liquid CDS markets in Asia.