Asian Domestic Currency Bond Markets

Local currency markets in East Asia show very limited integration with each other and highly variable integration with major global markets. A very limited exception to the generalization that East Asian fixed income markets are not integrated with each other is the issuance by the highest rated East Asian names in the Japanese yen bond markets. And, if the scope of the analysis is widened to include East Asia and the Pacific, there is extensive integration of the Japanese and Australian and New Zealand fixed income markets in the form of investment by Japanese investors in Australian and New Zealand dollar-denominated bonds. (At this stage of the exchange rate cycle of the Australian and New Zealand dollars, however, this integration is not altogether comfortable for the receiving countries.) Given the lack of integration in domestic currency bond markets in East Asia, the Asian Bond Fund 2 initiative represents an attempt by the authorities in the region to provide leadership to private market participants.

A. Local Bond Markets: Relatively Closed and Variably Responsive to Global Pricing

In contrast to the regionally integrated international bond and syndicated credit markets, most local currency bond markets in East Asia remain quite local affairs.⁸ Regional bond markets were characterized as unintegrated by Park and Bae (2002) and Eichengreen and Park (2005). More recently, however, analysis of the matrix of cross-border bond holdings has found evidence of Asian integration (Lee, 2006; Eichengreen and Luengnaruemitchai, 2006a), albeit, unsurprisingly, not as strong evidence as in the euro area.

In all likelihood, the regional integration identified in these recent studies reflects the pattern of ownership of US dollar bonds issued by Asian borrowers, as described above. Evidence for this supposition comes from the US data on foreign bond holdings, which are broken down by currency. US holdings of bonds issued by borrowers in East Asia were denominated in domestic currencies only to the extent of 15% of total holdings despite the fact that the stock of local currency debt far exceeds that of foreign currency debt across Asia (Table 1 [ PDF 88.3KB | 1 pages ]).⁹ Even if the US share of local bonds is lower than that of other holders of Asian bonds, most of the regional integration found in studies of the coordinated portfolio investment survey could reflect US dollar bond holdings.

Turning to the correlation of returns, the impression that Asian local currency bond markets are local affairs is reinforced (Figure 3 [ PDF 71.6KB | 1 pages ]). In the case of China and India, the low or even negative correlation in this sample period reflects the capital controls that effectively limited non-resident investment, as well as a lack of business cycle synchronization. In the case of Indonesia (not included) and the Philippines, idiosyncratic political developments worked in this sample period to lower the co-movement with global markets. The correlations in Korea; Taipei,China, and Thailand are moderate and reflect segmentation arising from withholding taxes and other impediments to cross-border investment (Takeuchi, 2006). Only Hong Kong's and Singapore's bond markets represent the exceptions, showing strong links with major markets as represented by the US Treasury market. (It would be worthwhile to test for co-movement of the idiosyncratic returns in Asian bond markets, with the prior expectation that such co-movement is weak.)

One way of looking at these very different degrees of integration into global bond markets is that the markets in East Asia are at different points along the path taken by the Australian bond market (Figure 4 [ PDF 71.6KB | 1 pages ], updating Kortian and O'Regan, 1996). This market went from insularity maintained by capital controls in the early 1980s, through low but highly variable correlation as non-residents began to invest in the domestic market and offshore issues, to the high integration of the late 1990s (McCauley, 2006b). These days most of the movement of bond yields happens between the close of Sydney and its opening the next day; the US employment report is the piece of news that moves the market most (Kearns, 2006). For years the close linkage could be ascribed to business cycle similarity, but this broke down in the early years of this century, while the linkage of the markets remained.

It should be noted, however, that Asian local currency bond markets might not end up as highly correlated with the US dollar bond market as the Australian dollar bond market, or the euro bond market for that matter. The yen bond market has retained a low correlation notwithstanding the deep involvement of foreign banks and securities firms in the market. Although there is some recent evidence of closer linkage of the yen bond market and global bond markets, the record from the mid-1980s into this century, at least, suggests that Asian local markets might not be on the path taken by the Australian bond market.

B. The Samurai Market

The yen bond market for East Asian issuers has not contributed as much to regional integration as one might have anticipated given Japan's status as a creditor country. There was a time before the Asian financial crisis when it seemed that Asian issuance in the yen bond market in Tokyo (the so-called Samurai market) might really take off and that this market might develop a strong regional bias. But, as noted by Ogawa (2005), Asian issuance in this market never really recovered from the Asian financial crisis (Figure 5 [ PDF 100.6KB | 1 pages ]).

Why is this so? Nishi and Vergus (2006) note that lack of recovery of Asian Samurai issuance reflected the current account and corporate surpluses in the region, as well as the stronger appeal of the euro bond market to regional issuers as compared with that of the predecessor currencies. Still, Asian issuance of US dollar bonds did recover its pre-crisis levels, so the performance of the yen market remains to be explained.

In retrospect, the timing of the elimination of the requirement for an investment grade rating for Samurai issuance in 1996 was unfortunate for the integration of Asian capital markets. This allowed issuance by an Indonesian name and a Hong Kong based securities firm with considerable credit exposure in Indonesia. Their defaults in 1997 left Japanese investors hesitant to buy Asian bonds issued by names other than the best sovereigns or quasi- sovereigns. This hesitation was reinforced by a subsequent default of a Chinese provincial name in 2000, the difficulties of Xerox, and the defaults of Enron in the euroyen market and Argentina in the Samurai and euroyen markets (Nishi and Vergus, 2006; Schmidt, 2004).¹⁰ The contemporaneous arrival of better-understood sub-investment grade domestic corporate issues allowed a home bias in credit risk. At the same time, Japanese investors drew the conclusion that they understood currency risk if not credit risk. This set the stage for the development of the Uridashi market not only for US dollar bonds but also those denominated in euro and Australian and New Zealand dollars. In short, a manifestation of credit risk at an early stage of Japanese investment into Asian credit all but cut off this channel of Asian capital market integration.¹¹

C. The Uridashi Market for Australian and New Zealand Dollar Bonds

As noted, the revulsion of Japanese investors to regional credit risk has been matched by their embrace of currency risk in the form of high-coupon foreign currency bonds. These have disproportionately included Australian and New Zealand dollar bonds.

A remarkable feature of this integration of Australian and New Zealand dollar bonds into the bonds regularly offered to Japanese investors, particularly households, is that the distributors, and not the issuers, have made the investment in investor education. The issuers act, in the main, purely opportunistically, using their high credit ratings to secure cheap funding as measured against US dollar Libor or euribor. In contrast, the distributors, chiefly the Japanese securities firms, in effect market Australia and New Zealand as well as their currencies to the investors. For instance, the head of the Japanese securities firm's affiliate in Sydney, rather than any issuer, conducts the investor seminars not only in Tokyo but around Japan. Housewives can pose questions about the outlook for interest and exchange rates in the Antipodes.

A question then for the financial integration of Asia then comes into view. Under what circumstances would it make sense for the Japanese securities firms to similarly “invest” in the marketing of an Asian economy and bonds denominated in its currency?

D. The Asian Bond Fund 2 Initiative

The Asian Bond Fund 2 followed a previous, less ambitious initiative by the same group of East Asian and Pacific central banks. In July 2003 they pooled $1 billion of their reserves in a fund that purchased US dollar-denominated government and quasi-government bonds issued by eight of the eleven economies involved. If one accepts the argument above that East Asian investors figure prominently among the buyers of US dollar-denominated bonds sold by East Asian issuers, then the first Asian Bond Fund (ABF1) allowed the central banks to catch up with private portfolio managers.¹²

In contrast, the Asian Bond Fund 2 (ABF2) initiative of the Executive's Meeting of East Asia Pacific central banks (EMEAP) can be seen as an effort by officials to lead the markets. For one thing, the initiative flagged a new asset class for global private investors. For another, East Asian investors were given the example of official reserve managers' investing in a low- cost indexed product.

If the ABF2 sought to lead market participants, then their response to it merits some analysis.¹³ The $2 billion invested by the central banks was intended to serve as no more than seed money. The aim was to catalyze private investment in bond funds managed against indices created and maintained by a global index provider. While to some extent the value of the project lay in the reforms induced by the process and the force of comparison (EMEAP, 2006; Ma and Remolona, 2005) as well as the habits of cooperation engendered, the private investment drawn to the project offers one metric for the results of the effort.

In assessing the private response to the ABF2, it must be recalled that the opening of the various funds to the public occurred unevenly, in part reflecting the challenges posed by the introduction of an innovative product. Thus, while the Pan-Asia Index Fund was launched in July 2005, individual market funds were subsequently launched over the next nine months in Hong Kong, Korea, Malaysia, Singapore; and Thailand. Then over the period April 2006– March 2007, further public funds were offered in Indonesia and the Philippines. At writing, only the individual market fund for China remains to be offered to the public (Table 2 [ PDF 69.4KB | 1 pages ]).

In the 21 months between the Pan-Asia Index Fund's launch on the Hong Kong Stock Exchange and end-March 2007, its 17% total return has drawn private investment. This has helped to boost its size to more than half again larger than the original central bank investment. If the 19% growth of the PAIF over the original EMEAP listing through April 2006 was judged “satisfactory” and comparable to that of other bond funds in the region (EMEAP Working Group on Financial Markets, 2006: 3), that 55% growth through March 2007 seems quite respectable.

The PAIF is the ABF2 fund that serves most directly to increase the integration of capital markets in Asia. As it expands, larger sums of local currency bonds are gathered into a single portfolio. Moreover, the origin of the investors other than central banks has also contributed to financial integration. It is said that Japanese institutional investors have been prominent among the investors. A yen-denominated feeder fund for the PAIF has been established in Japan to centralize the currency conversion from yen to US dollars (EMEAP Working Group on Financial Markets, 2006: 22).

The individual funds have also drawn private investors to varying extents. Taking the individual funds as a whole, investment by private investors along with price increases in some cases have boosted the sum of the funds under management to, again, more than half again the scale of the original central bank seed money.

All in all, the ABF2 assets under management have grown from the initial central bank investment of $2 billion to about $3.1 billion. Of this growth, $800 million represents the private sums drawn to the funds. Private funds invested in the ABF2 family of funds could surpass that of the central banks in 2008.

These results should be evaluated against the backdrop of the relative underdevelopment of exchange-traded funds (ETFs) in Asia. These low-fee, listed index products were only introduced into Asia with the sale of the Tracker Fund by the Hong Kong authorities in 2000 (Ho, 2006). By mid-2006, the number of ETFs in Asia had grown to 59, with $53 billion under management, mostly in Japan. These numbers lagged well behind the equivalent numbers from the United States and even from Europe (Table 3 [ PDF 65.4KB | 1 pages ]). The diffusion of this financial innovation in Asia has been slowed not only by the Asian preference for safety in bank deposits, on the one hand, and high-risk, high-return investments, on the other (initial public offerings, structured products). In addition, commission-based distribution channels face no incentives to sell these low-fee products. As the EMEAP Working Group on Financial Markets (2006: 21–2) put it,

The initiative has succeeded in bringing an alternative low-cost vehicle relative to what is currently available in each of the eight markets. However, the low-cost structure is a double-edged sword. While a low expense ratio is an attractive feature to investors, it limits the resources available for providing the necessary incentives for distributors to promote this product to the retail investors.

In the United States, financial planners, who, unlike banks or brokers, do not take commissions for the sale of funds to their principals, steer investors to low-cost products like exchange-traded funds. Such planners do not yet figure prominently in Asian investing circles.

Thus, the PAIF for now relies on its own performance for its advertisement. The low fees help lift the PAIF on the league tables that are now compiled by the mutual fund performance tracker Lipper and published in the financial press. (This compilation and publication themselves testify to the acceptance by market participants of a new asset class.) Low fees are no guarantee of relative performance if unindexed private managers get it right, at least for a time, and thereby earn their fees. But the PAIF ran well right out of the box.

Exhibit 1: “Fund Scorecard: Asia Bond Funds” [ PDF 251.3KB | 1 pages ]

Exhibit 2: “Fund Scorecard: Asia Bond Funds” [ PDF 188.5KB | 1 pages ]

E. Summing Up

In sum, fixed income markets for Asian issuers stand at two extremes. At one end is a well- integrated US dollar bond market that features East Asian buyers of East Asian bonds to a considerable extent. At the other extreme are the local currency bond markets of Asia that remain very local affairs. That internationally syndicated loan markets in some local currencies also bring together Asian borrowers and lenders point to the possibility of more integrated domestic currency bond markets. For now, the central banks in the region have tried to narrow the gap between these extremes. They have helped to define Asian local currency bonds as an asset class and led by example by working with market participants to set up cost-effective vehicles for gaining exposure to this asset class.

Download this Discussion Paper [ PDF 550.5KB| 27 pages ].

Comment(s)

There are [0] comment(s) for this entry. Post a comment.