Abstract

Private market financial flows to developing countries increased significantly in 1992 and 1993 (Chart 4). Bond and equity flows accounted for much of the increase. The rapid expansion was mirrored in a broadening of the range of developing country borrowers attracting international investors, although portfolio flows continued to be concentrated in a few key countries in Asia and Latin America. In contrast, medium-and long-term bank lending to developing countries remained moderate; banks did demonstrate renewed interest in such lending, but on a highly selective basis.

Private market financial flows to developing countries increased significantly in 1992 and 1993 (Chart 4). Bond and equity flows accounted for much of the increase. The rapid expansion was mirrored in a broadening of the range of developing country borrowers attracting international investors, although portfolio flows continued to be concentrated in a few key countries in Asia and Latin America. In contrast, medium-and long-term bank lending to developing countries remained moderate; banks did demonstrate renewed interest in such lending, but on a highly selective basis.

Chart 4.
Chart 4.

Private Market Financing to Developing Countries

(In millions of U.S. dollars)

Sources: International Financing Review; OECD; and IMF staff estimates.1Medium- and long-term bank loan commitments only.

In 1994, the situation changed dramatically. With higher U.S. interest rates, as well as unfavorable economic and political developments in some major borrowing countries, bond and equity issuance by developing countries plummeted between February and April 1994. A modest recovery came in the following months, but new flows remained vulnerable, especially because of the uncertain course of U.S. interest rates. Despite the market correction, however, portfolio flows to developing countries in the first half of 1994 were still significantly higher than levels recorded in the early 1990s.

Bonds

Bond placements by developing country borrowers reached $59.4 billion in 1993, more than twice the amount placed in 1992 (Tables 5 and A5).11 There was a strong acceleration in bond issuance in the final quarter of 1993; bonds issued in that quarter amounted to $23.7 billion, almost equal to total issuance activity in 1992. This surge reflected a decline in U.S. interest rates combined with relatively high returns in emerging markets. Both factors encouraged a broader range of mainstream institutional investors to participate more actively in these markets. The continued implementation of prudent macroeconomic policies and structural adjustments in borrowing countries also improved investors’ confidence. In relative terms, developing countries continued to increase their share of total international bond issuance from 7.1 percent in 1992 to 12.4 percent in 1993, and to 20.1 percent in the fourth quarter of 1993. The average size of developing country bond placements also increased from $111 million in 1992 to $125 million in the first half of 1993, and to $135 million in the second half of the year. In particular, there were a number of sizable issues by borrowers in Latin America.

Table 5.

International Bond Issues by Developing Countries and Regions1

(In millions of U.S. dollars)

article image
Sources: IMF staff estimates based on Euroweek, Financial Times, International Financing Review, Financial Market Trends, Financial Statistics Monthly, and Organization for Economic Cooperation and Development.

Including note issues under Euro medium-term notes (EMTN) programs.

In the first half of 1994, volatile market conditions led to a sharp decline in the volume of international bond issuance by developing countries. Bonds worth $26.1 billion were issued, most in the beginning of the year and in June. Beginning in February, as bond yields rose throughout the world, both issuers and investors pulled back. For the first half of 1994 as a whole, the developing country share of total international bond issues fell to 12 percent. The decline was particularly notable for countries in Europe and Latin America.

The terms on new issues for many developing country borrowers improved throughout 1993 and into early 1994. The average yield spread fell from 288 basis points in the first quarter of 1993 to a low of 187 basis points in the first quarter of 1994 (Chart 5).12 The average spread widened sharply in the second quarter of 1994 to 259 basis points. This increase occurred despite the fact that bonds during this period were issued primarily by borrowers with higher credit ratings and carried shorter maturities. The weighted average maturity of bonds issued shortened to 6.3 years in the first half of 1994, after lengthening from 5.4 years in the first quarter of 1993 to 7.2 years in the final quarter of the year. Yield spreads in the secondary market for developing country bonds followed a similar pattern; notable increases were observed for bonds issued by Argentina, Brazil, Hungary, Mexico, Turkey, and Venezuela (Chart 6). A tightening of market conditions was also reflected in a sharp decline in secondary market prices for Brady bonds.

Chart 5.
Chart 5.

Yield Spreads at Launch for Unenhanced Bond Issues by Developing Countries1

(In basis points)

Sources: International Financing Review; and Financial Times.1Reflect weighted averages.
Chart 6.
Chart 6.

Secondary Market Yield Spreads on U.S. Dollar Denominated Bonds by Selected Developing Countries

(In basis points)

Source: Reuters.

Yield spreads continued to vary considerably among countries. While borrowers without records of debt-servicing difficulties commanded lower spreads, market re-entrants were typically faced with spreads of over 200 basis points in 1993 and the first half of 1994. Moreover, the spread for public sector borrowers continued to be substantially lower than that for private sector borrowers (Table A6). Reflecting the uncertain path of U.S. interest rates, there was also renewed interest in new issues bearing floating rates. In the first half of 1994, floating rate notes accounted for some 20 percent of total new issues, up from 8 percent in 1993.

Most of the recent bond issuance by developing countries represents net capital inflows. Maturing developing country bonds amounted to only $6.3 billion in the period from 1991 to 1993, compared with total bond issues worth $96 billion. As of the end of June 1994, the total outstanding stock of international bonds issued by developing countries is estimated to stand at $117.5 billion, 42 percent of which was accounted for by private sector borrowers. Amortization payments on this stock of debt will rise sharply in the next few years from an estimated $7.1 billion in 1994 to a peak of $21 billion in 1998, as bullet repayments on bonds placed in the early 1990s fall due (Table A7).13 Payments are particularly concentrated in a few Western Hemisphere and Asian countries.

Although the range of borrowers continued to widen in 1993, developing country bond issuers remained concentrated in a few countries in Latin America, Asia, and Europe. Latin American borrowers had the largest share until they were surpassed by Asian borrowers in the final quarter of 1993. Bond issues by Latin American borrowers doubled to $27.4 billion in 1993, accounting for 46 percent of total issues by developing countries. That figure declined in the first half of 1994 to $8.9 billion, or 34 percent of total issues by developing countries.

Mexico continued to be the leading borrower, raising $10.8 billion in 1993 and $4.7 billion in the first half of 1994. An increasing number of companies placed bonds, including several private and public banks. Moreover, a few firms floated quite sizable individual issues during this period, including $1 billion issues by Cemex (Mexico’s largest cement producer) in May 1993 and Bancomext in January 1994. Mexican entities also opened up new currency sectors, with Mexico’s launching of Latin America’s first “Samurai” issue since the debt crisis, Banamex issuing a Mexican peso-denominated Eurobond, and Nafinsa (a Mexican development bank) placing a “Dragon” bond, marking the first time that a Latin American noninvestment grade borrower has been able to issue in that market.

Borrowers in Argentina quadrupled their bond issues to $6.2 billion in 1993 and raised $2.4 billion in the first half of 1994. There were a number of new developments. A $1 billion global sovereign bond was issued in December 1993 at a spread of 280 basis points, the first global bond ever issued by a developing country borrower.14 In addition, the first significant convertible bond by a private telephone company based in a developing country was placed in March 1994. Brazil also increased its bond issuance significantly in 1993, notwithstanding continued uncertainty about the course of its economic policies. All Brazilian bonds were issued by nonsovereign borrowers and included the first Euro-yen issue by a Latin American entity at a spread of 416 basis points. Venezuelan entities also increased their borrowing activity in 1993; among others, a $1 billion bond was issued by PDV America (a state oil company) at a spread of 210–218 basis points, and the first intraregional bond was issued simultaneously in Colombia and Luxembourg. Mainly reflecting the concerns of investors about the country’s economic situation; however, no international bonds were placed by Venezuelan borrowers during the first half of 1994.

In general, the range of Latin American borrowers in the international bond market continued to broaden. Colombia, Guatemala, and Peru, and more recently Barbados, Bolivia, and Costa Rica, tapped the market for the first time in many years. Spreads ranged from 215 basis points for Colombia to over 700 basis points for Peru. At the same time, recent market re-entrants such as Chile, Trinidad and Tobago, and Uruguay maintained their presence in the market.

Asian borrowers tripled their international bond issues to $20.4 billion in 1993. They also raised $13 billion through this channel in the first half of 1994, despite overall market turbulence. As a result, their share in the total bond issues by developing countries rose from 34 percent in 1993 to 50 percent in the first half of 1994. The stock market booms in the region in 1993 also led to a strong increase in convertible bond issues; of total bonds issued, the share of convertibles rose from 18 percent in 1992 to 30 percent in 1993 before falling in the first half of 1994. Hong Kong emerged as the leading borrower in Asia, followed by Korea. Together, these two countries accounted for more than half of the bonds placed by Asian entities in 1993. The People’s Republic of China also significantly increased its recourse to the international bond market. To facilitate market entry for Chinese enterprises, after a six-year absence, the Government entered the market directly in 1993 and placed three issues intended to establish a benchmark for China risk. In February 1994, the Government issued a 10-year, $1 billion global bond, which was priced at an 85 basis point spread; the issue was predominantly purchased by U.S. investors. In 1993, India, Malaysia, and the Philippines also tapped the international bond market for the first time in several years. Pakistan and Macao entered the market in the first half of 1994. Concerned about the country’s overall debt profile, however, the Indian authorities moved in May 1994 to restrict convertible bond issues, except for companies using proceeds to restructure existing external debt.

European developing countries continued to step up their recourse to the international bond market in 1993, where they raised $9.6 billion. In the first half of 1994, however, bond issuance activity dropped dramatically, largely reflecting reduced placements by Hungary and Turkey. With increasing market concerns about economic conditions in these two countries, bond issues by Hungary fell from $4.8 billion in 1993 to less than $0.3 billion in the first half of 1994, while issues by Turkey declined from $3.9 billion to $0.8 billion. The Czech Republic and the Slovak Republic maintained access to the market in the first half of 1994.

In the rest of the developing countries, only a handful of borrowers have tapped the international bond market. Israel raised $2 billion in 1993 and another $2 billion in the first half of 1994 on exceptionally favorable terms because of guarantees provided by the U.S. Agency for International Development. In Africa, the Congo launched a $600 million ten-year bond, with interest payments secured by oil receivables and principal collateralized by U.S. Treasury bonds. In addition, the Central Bank of Tunisia issued the first Samurai bond by an African country; the bond was launched with a spread of 221 basis points above the yield of a risk-free yen bond of comparable maturity.

Among developing country bond issuers, private sector borrowers scaled back their international issuance activity in the first half of 1994, following the doubling in volume that occurred in 1993. As a result, their share in total bonds issued by developing country borrowers declined from 45 percent in 1993 to 26 percent in the first half of 1994. The lower issuance activity was partly due to the sharp contraction in convertible bond issues by entities in Hong Kong, which were adversely affected by a decline in stock prices. Bond issues by sovereign borrowers, which tripled in 1993, declined in the first half of 1994, but their share of total bond issues rose from 27 percent to 50 percent. Lower issuance by sovereign borrowers was mainly accounted for by Hungary and Turkey.

Most bonds issued by developing countries continued to be denominated in U.S. dollars, yen, and deutsche mark. Bond issues in U.S. dollars accounted for 74 percent of the total in 1993 and 81 percent in the first half of 1994 (Table A8). This high share partly reflected both the greater appetite of U.S. investors for high-yielding, subinvestment grade securities and the impact of relatively low U.S. interest rates. Also facilitating bond issues in the U.S. market is Rule 144a, which exempts private placements from the disclosure requirements of the Securities and Exchange Commission (SEC) and permits qualified institutional buyers to trade privately placed securities without waiting the usually stipulated two-year holding period. In 1993, several borrowers in Hungary, Korea, and Mexico tapped the Yankee bond market for the first time.15 While the yen sector remained the second largest currency sector for bonds issued by developing country borrowers, its share declined from 13 percent in 1993 to 10 percent in the first half of 1994. Deutsche mark bond issues, principally by European and Latin American borrowers, were subdued in 1993 and in the first half of 1994.

Following the market turbulence in early 1994, developing country borrowers increased the use of credit enhancement techniques, such as bond-equity conversion options, collateralization, and put options (Table A9). The most widely used enhancement technique was the bond-equity conversion option, especially in Asia where convertible bonds accounted for more than half of bond issues in the first half of 1994. Put options were the second most widely used technique; Latin American issuers were the principal users, possibly reflecting uncertainties about economic prospects in the region.

The expansion of the investor base was accompanied by an increase in the number of countries assigned credit ratings by major rating agencies (Table 6). During 1993 and the first half of 1994, Argentina, the Philippines, the Slovak Republic, Trinidad and Tobago, and Uruguay received initial subinvestment-grade ratings from the major U.S. credit rating agencies. Initial investment-grade ratings were also assigned to Colombia and Taiwan Province of China. In addition, Chile received an investment-grade rating from Moody’s in February 1994, making it the only Latin American country to have received such a rating by the two major U.S. rating agencies. In March 1993, the Czech Republic became the only developing country in Europe with an investment-grade rating, and this rating subsequently was upgraded in May 1994. Other investment-grade countries receiving upgraded ratings during 1993 and the first half of 1994 included Chile, China, Israel, Malaysia, and Singapore. Owing to deteriorating economic conditions, conversely, Turkey was downgraded to a subinvestment grade rating in January 1994, and its rating was reduced further a few months later. The major rating agencies also downgraded Venezuela in March and April 1994.

Table 6.

Credit Ratings of Developing Country Borrowers1

article image
article image
Sources: Financial Times; International Financing Review; and Salomon Brothers.

Ranked in descending order according to rating. Ratings by Standard and Poor’s and Moody’s Investor Service. The ratings are ranked from highest to lowest as follows:

In addition, numbers from 1 (highest) to 3 are often attached to differentiate borrowers within a given grade.

Bond Pricing

The rapid increase in the prices of developing country bonds during 1993 raises some questions as to whether the markets were adequately pricing the risk of these securities. Moreover, the markets have at times appeared to react slowly to changes in economic conditions in a country.16 In pricing bonds, the markets are reported to use two basic approaches. Risk may be priced on a relative basis, with bond yields set in relation to some benchmark issue or to the securities of other issuers judged to be of roughly comparable risk. Alternatively, market participants may attempt directly to assess risk using a scoring system based on a set of economic and political factors. Such scoring systems may vary widely in terms of their level of quantification, the factors considered, and the relative importance assigned to individual factors over time. All involve a high degree of judgment.

Observers suggest that Mexican bonds are frequently used as a benchmark, since Mexico is generally viewed as one of the best credit risks among those developing countries that previously rescheduled debts to private foreign creditors. To test this proposition, Granger causality tests were run on daily bond prices for sovereign issues by Argentina, Brazil, Hungary, Nigeria, the Philippines, Turkey, and Venezuela. The tests used the price of the Mexican par bond as the benchmark for Brady bonds and the price of a Mexican new issue as the benchmark for Eurobonds. The results suggest that movements in the price of the Mexican par bond precede price movements in the par bonds of Argentina, Brazil, and the Philippines (Table A10). The price movements of the Mexican par bond, however, did not precede movements in either the prices of the Venezuelan or Nigerian par bonds, a result that could reflect country-specific factors.17 Price movements in the Mexican Eurobonds were found to precede changes in the prices of most of the Eurobonds issued by the other countries sampled (Table A11).

The pricing of a bond should be in line with that of other bonds considered by the market as roughly comparable in terms of risk. The relationship between prices on U.S. corporate bonds and on bond issues from developing country sovereign borrowers with the same credit rating can provide some indication of how the markets view developing country risk (Chart 7). Deviations in the pricing of the two types of bonds, however, represent either mispricing of the riskiness of developing country bonds or market perceptions that such bonds are in fact riskier than their U.S. corporate counterparts (i.e., in the market’s view, the rating agencies’ ratings of developing country bonds are not accurate). IMF staff analysis indicates that Mexican and Philippine par bond prices and Hungarian and Turkish Eurobond prices do roughly track the prices of comparably rated U.S. corporate bonds, although prices for the latter two bonds are more volatile.18 Argentina appears to have undergone a sharp reappraisal by the market of its creditworthiness in late 1992, and its yield steadily approached that of the comparable U.S. corporates, until early 1994. In the case of Venezuela, market perceptions of its creditworthiness may have adjusted much faster than the country’s credit rating. The markets appear to have incorporated another downgrading into their pricing of Venezuelan bonds after the major credit rating agencies placed Venezuela on a credit watch. The market also appears to consider the unrated Brazilian and Nigerian par bonds as the equivalent of bonds rated below Caa.

Chart 7.
Chart 7.

Comparison of Yields of Sovereign Bonds with Yields on U.S. Corporate Bonds

(In percentage points)

Sources: Moody’s; Reuters; and Salomon Brothers.1Venezuela placed on watch list by Moody’s.2Venezuela downgraded by Moody’s to Ba3.3Turkey placed on watch list.4Turkey downgraded by Moody’s to Bal.5Turkey downgraded by Moody’s to Ba3.

The scoring systems used by market participants attempt more systematically to consider country-specific factors in bond pricing. Countries are ranked on the basis of a number of political and economic variables, and these rankings are used to assess the spread between the yield on a developing country bond and that on a “risk-free” bond. Factors often considered include (1) political conditions (e.g., the government’s commitment to economic reform, its ability to implement policies, and popular support); (2) macro-economic conditions (especially inflation, growth prospects, and fiscal policy); (3) structural reform; and (4) the country’s balance of payments position and prospects. Data for the period 1989 to 1994 generally support the view that the market applies an ordinal ranking of developing country sovereign bonds in line with underlying economic fundamentals.19 Among the Latin American Brady par bonds, Mexico carries the lowest spread, followed by Argentina, Venezuela, and Brazil (Chart 8). The markets considered Venezuela a better risk than Mexico prior to 1991, but since then Venezuela’s spread has increased to approach that of Brazil. In contrast, the premium above Mexico paid by Argentina has declined steadily. The Philippines has also paid a spread higher than Mexico, and Nigeria’s spread has been above that of the Philippines. Over time, though, the Philippine premium has narrowed, while Nigeria’s spread has widened, reflecting differences in the relative economic performance of the two countries. Data from representative Eurobond issues present a more or less similar picture, taking into account the different duration and liquidity characteristics of the bonds.20

Chart 8.
Chart 8.

Comparison of Sovereign Bond Spreads

(In basis points)

Sources: Reuters; Salomon Brothers; and IMF staff estimates.

The movement in the secondary market spreads over time suggests a rough relationship with a country’s rate of inflation and its level of foreign assets, perhaps because of the frequency of the availability of this information. For example, a very simple examination of the data reveals that the steady downward trend in the spread on Mexican bonds appears to have coincided with a period of declining inflation and rising foreign assets (Chart 9). The recent widening of the spread on Venezuela’s bonds roughly coincided with a drop in net foreign assets and lagged a deterioration in inflation performance. The spread on Turkish bonds fell to less than 200 basis points by February 1994, before rising precipitously in March 1994; this turnaround took place five months after gross official reserves began to fall but roughly coincided with a reported surge in inflation (Chart 10). In the case of Hungary, the bond spread fell by roughly 200 basis points between September and December 1993; it then began to increase in early 1994, as foreign reserve assets declined and inflation turned up.

Chart 9.
Chart 9.

Comparison of Movements in Spreads and Economic Variables for Mexico and Venezuela

Sources: Reuters; Salomon Brothers; and IMF staff estimates.
Chart 10.
Chart 10.

Comparison of Movements in Spreads and Economic Variables for Hungary and Turkey

Sources: Reuters; Salomon Brothers; and IMF staff estimates.

Equities

In contrast to bonds, the growth of equity placements in the international capital market moderated in 1993, after expanding sevenfold between 1990 and 1992 (Table 7). Issuance activity did pick up, however, in the final quarter of 1993, reflecting buoyant stock markets in Asia and Latin America. International equity placements by developing countries increased by 28 percent to $11.9 billion in 1993, but their share in total international equity placements declined to 23 percent from 41 percent in 1992. In the first half of 1994, developing country equity issues declined only moderately to $6.9 million from $7.7 billion in the second half of 1993. The lower issuance activity was accompanied by a general decline in share prices in major developing country stock markets.

Table 7.

International Equity Issues by Developing Countries and Regions

(In millions of U.S. dollars)

article image
Sources: IMF staff estimates based on Euroweek, Financial Times, International Financing Review (IFR), and IFR Equibase.

Most international equity placements have been accounted for by Latin American and Asian companies. Latin American companies raised $5.7 billion in the international equity market in 1993 and $2.1 billion in the first half of 1994. Argentina emerged as the leading Latin American issuer, with issues increasing from $0.4 billion in 1992 to $2.8 billion in 1993. Issues occurred mainly through the vehicle of American depositary receipts (ADRs)21 and were predominantly accounted for by the privatization of Yacimientos Petroliferos Fiscales (a state-owned oil and gas company) which raised $2.4 billion. Equity issues by Mexican companies declined from $3.1 billion in 1992 to $2.5 billion in 1993, partly because of uncertainty over approval of the North American Free Trade Agreement. Following approval, Mexican companies raised $1.7 billion in the final quarter of the year, mainly through ADR and global depositary receipt (GDR) programs;22 this included an $822 million GDR offering by Grupo Televisa. Companies in Bolivia, Colombia, and Peru entered the market for the first time in 1993, together raising $127 million.

Asian companies raised $5.7 billion in 1993 and $3.8 billion in the first half of 1994. In 1993, companies from China were the leading issuers, raising $1.9 billion, or double the amount raised in 1992. In addition to “B” shares listed in Shanghai and Shenzhen and reserved for foreign investors, Chinese companies in July 1993 began issuing shares listed on the Hong Kong Stock Exchange, so-called H shares. Other major developments included the first global equity placement by a Chinese company in July 1993 that combined ADRs and H shares, and the first equity placement in August 1993 by a private Chinese company on the Hong Kong Stock Exchange. Indian companies stepped up equity placements, raising over $1 billion in early 1994, compared with $0.3 billion in 1993, as they shifted away from convertible bonds to take advantage of the premium provided by a stock market boom. In light of the subsequent weakness in markets worldwide, however, Indian equity placements declined significantly. Indonesian entities, for their part, accelerated their equity issuance until March 1994, when the authorities moved to stem capital inflows. During the same period, a company in Sri Lanka entered the international equity market for the first time through GDRs, and entities from Bangladesh tapped the international equity market through small issues denominated in local currency.

International equity issues by companies in the rest of the developing world remained limited, with the exception of Turkey and Ghana. In the first half of 1994, a Turkish automobile company (TOFAS) raised $330 million through ADRs and GDRs. Ashanti Gold-fields (a gold-mining company) in Ghana raised $398 million through GDRs. Companies from Morocco and Poland entered the market for the first time in 1993 with small issues.

Over the past few years, direct equity purchases by international investors on local exchanges have become another important source of equity inflows for several developing countries. Although comprehensive statistics are not available, fragmentary information suggests that the direct purchases of equities have been quite sizable. In addition, mutual funds have become increasingly important sources of equity flows to developing countries. The number of so-called emerging market mutual equity funds increased from 91 in 1988 to 465 in 1992 and 573 in 1993. Total net assets in these funds rose from about $6.0 billion in 1988 to $81.5 billion in 1993 (Chart 11 and Table A12).23 The significant rise in the net asset position of emerging market mutual funds in 1993 largely reflected a sharp run-up in share prices in developing countries. The overall price index compiled by the International Finance Corporation (IFC) for developing country stocks that foreign investors are allowed to purchase (referred to as the investable index) rose by 75 percent in 1993 (Charts 12 and 13). Share prices in some Asian markets more than doubled during that year. Adjusting for these share-price increases, net purchases of developing country equities by emerging market mutual funds (including purchases of equities issued in international capital markets) can be approximated.24 Mutual fund purchases of developing country equities are estimated to be $12.6 billion in 1993, compared with $8.4 billion in 1992 (Table A13). Issuance of closed end emerging market mutual funds accounted for $2.0 billion of the total in 1993 (Table A14).

Chart 11.
Chart 11.

Emerging Market Mutual Funds

(In billions of U.S. dollars)

Sources: Emerging Market Funds Research, Inc; and Lipper Analytical Services, Inc.1Net flows to developing countries are estimated by deflating changes in net assets of funds by IFC investable share price indices.
Chart 12.
Chart 12.

Share Price Indices for Selected Markets In Latin America

(IFC Weekly Investable Price Indices, December 1988 = 100; in U.S. dollars)

Source: IFC Emerging Markets Data Base.1Argentina, Brazil, Chile, Colombia, Mexico, and Venezuela.
Chart 13.
Chart 13.

Share Price Indices for Selected Markets In Asia

(IFC Weekly Investable Price Indices, December 1988 = 100; in U.S. dollars)

Source: IFC Emerging Markets Data Base.1India, Indonesia, Korea, Malaysia, Pakistan, the Philippines, Taiwan Province of China, and Thailand.

Despite a sharp drop in stock prices, the number of emerging market mutual equity funds increased by more than 90 in the first quarter of 1994, probably reflecting the time lag involved in establishing these funds. The net asset value of all funds increased by about $9 billion, despite the large decline in share prices that occurred during that quarter. Data on open-end emerging market mutual funds domiciled outside the United States, however, suggest that the growth of emerging markets funds slowed considerably in the second quarter of the year. Issuance of shares in closed end emerging markets mutual funds also declined sharply from $4.2 billion in the first quarter of 1994 to $0.5 billion in the second quarter.

Mutual funds targeting Asian developing countries accounted for over 50 percent of total net assets in emerging market funds during 1993. Although global mutual funds have recently expanded considerably, it is reported that their investments have also been concentrated on Asian equities. Mutual funds designated for Latin American countries accounted for only 12 percent of the total net assets in emerging market funds.

Commercial Bank Lending

Banks began to show a renewed interest in lending to developing countries in 1993. In contrast to portfolio flows, however, the increase in medium- and long-term bank lending was modest. Medium- and long-term bank commitments to developing countries increased by 7 percent to $21 billion during the year (Table A15).25 Banks in general, however, shortened maturities, raised spreads, and continued to use a variety of risk-reducing techniques like asset securitization. The weighted average maturity of uninsured bank credits to developing countries declined from 6.7 years in 1992 to 5.5 years in 1993 (Table 8) and the weighted average spread over the LIBOR rose from 86 basis points in 1992 to 106 basis points in 1993. In the first half of 1994, the weighted average maturity increased to 6.8 years, and the spread narrowed to 99 basis points. But actual spreads varied considerably among developing countries ranging from 60 to 70 basis points for borrowers in Korea and Malaysia to 300 basis points for borrowers in India and Mexico (Table A16). The U.S. dollar continued to be the most important currency of denomination, accounting for over 80 percent of total syndicated loans to developing countries in 1993.

Table 8.

Terms of Long-Term Bank Credit Commitments1

article image
Sources: Organization for Economic Cooperation and Development (OECD), Financial Market Trends; and IMF, International Financial Statistics (for Eurodollar and prime rates).

The country classification and loan coverage are those used by the OECD.

The bulk of syndicated bank loans continued to be directed toward Asia, although lending to Latin America increased significantly in 1993. After having declined in 1992, uninsured medium- and long-term bank loan commitments to Asian borrowers increased to $15.7 billion in 1993. China continued to be the largest Asian borrower, receiving loans totaling $3.6 billion in 1993. It was followed by Thailand ($3.4 billion), Hong Kong ($2.0 billion), Korea and Indonesia ($1.9 billion each), and Malaysia ($1.6 billion). In the first half of 1994, loan commitments to Asia amounted to $9.2 billion, including $3.7 billion to Thailand. Other noteworthy developments included the first sovereign loan to Indonesia since 1991 ($400 million) and a $1.2 billion loan to a petroleum company in Thailand, the largest single borrowing in the Asian market in the past five years.

In Latin America, uninsured medium- and long-term bank loan commitments increased from $0.9 billion in 1992 to $2.2 billion in 1993. Venezuela increased bank borrowing to $0.8 billion in 1993, partly owing to loans raised by public sector oil exporters. Argentina received $0.4 billion in commitments for the first time in the 1990s, while Brazil, Chile, and Mexico maintained access to bank credits on the order of $0.2-0.4 billion. In the first half of 1994, however, bank loan commitments to Latin American borrowers declined sharply to only $0.2 billion.

New commitments to developing countries in Europe remained subdued, amounting to $2.6 billion in 1993 and $0.6 billion in the first half of 1994. Those commitments were confined to a handful of countries. Turkey continued to be the major borrower, receiving new commitments of $1.9 billion in 1993; however, banks in 1994 became more cautious in light of the country’s economic difficulties. Bank loan commitments to other European countries remained small, although several countries made their debut in the international credit markets, including the Czech Republic and Slovenia. A widely publicized DM 1.4 billion cofinancing facility for the Czech Republic’s Skoda Automodilova was canceled in September 1993, as its German parent revised its international investment activity. New bank commitments to the Middle East increased to $1.3 billion in the first half of 1994, compared with $0.4 billion in 1993, while bank lending to Africa remained almost nonexistent.

Other Issues

The strong growth in demand for developing country securities in 1993 occurred even though these assets were generally riskier than their counterparts in developed financial markets. Returns on equities in some developing countries have been significantly more volatile than those in the United States. Similarly, returns on selected developing country bonds have tended to be more volatile than returns on comparable U.S. Treasury bonds (Chart 14). Despite the higher risk inherent in developing country equities, investors sought the significantly higher returns offered by these assets and increasingly perceived that these would offer good opportunities for risk diversification.

Chart 14.
Chart 14.

Weekly Volatility of Total Returns on Bonds and Equities for Selected Countries1

(In percent)

Sources: International Finance Corporation; and Reuters.1Simple volatility averages of representative bond issues and IFC investable equity indices for Argentina, Brazil, Hungary, Korea, Mexico, Thailand, Turkey, and Venezuela.

The evidence shows that equity returns in many developing countries have low or negative correlations with equity returns in the United States and other developed financial markets and that returns among many developing country equities are relatively uncorrected.26 This pattern of correlations suggests that the expected return of a portfolio can be increased for a given level of risk by adding developing country equities, even though these assets are riskier on average than equities in industrial country markets.27 Total returns on developing country bonds (both Brady bonds and new issues) have also been relatively uncorrelated with the total return on comparable U.S. Treasury bonds and among themselves (Tables A17 and A18).

These historical correlations change over time and are not a certain guide to future relationships between returns on industrial and developing country securities. The correlation between prices on Brady bonds and U.S. Treasury securities rose substantially in 1993 and the first half of 1994. Returns on new developing country bond issues also became more highly correlated with U.S. bond returns and among each other in the first half of 1994. Correlations among equity returns appear to have been more stable, although for some countries correlations rose significantly in the first half of 1994 (Table A19). This evidence suggests that the benefits to be gained by diversifying into developing country securities may be weaker in periods of significant market disturbances.

The volatility of returns on many developing country securities has not tended to diminish over time, and for many countries it increased in the first half of 1994. Strong movements in 1994 followed a period of rapid expansion in the investor base, which raises questions about the relationship between market volatility and the stability of that base. In the traditional analytical view, growth in the number of investors should diminish volatility over time as new investors add liquidity and more diversity of risk preferences. This implies that a market could be cleared with smaller movements in prices.28 Such a view, however, is based on the assumption that differences in expectations, risk preferences, and liquidity needs among investors are purely random, which means that new investors are essentially drawn from the same population as existing investors. If, instead, new investors differ systematically from existing investors with respect to such considerations, an expansion of the investor base could contribute to higher volatility. For example, if new investors have much poorer information, their expectations will be more variable, and their entry into the market could add to price volatility.

The traditional framework also assumes that no individual investor is able to influence the price of assets traded in the market. Recent analytical work, however, has focused on the implications of a less-competitive structure in financial markets.29 In such a context, the investor base in a market can be assumed to consist of small, risk-averse investors, arbitragers, and a few large investors with inside information and an ability to influence market prices. To maintain the value of their inside information, large investors might try to conceal their trades from other market participants, suggesting that the price of the asset will not necessarily reflect all available information. In this situation, an increase in the number of large investors could at least initially add to volatility.30

The experiences of developing countries that have opened their stock markets to foreign investors do not provide a clear picture as to the relationship between market volatility and the investor base. The volatility of equity returns in 17 developing countries was compared before and after the opening of their equity markets to foreign investors; the results show that price volatility increased in eight cases (Table 9).31 The volatility of equity purchases was also examined for 14 of these countries using U.S. balance of payments data that provide country details on purchases by U.S. residents.32 In all cases, these data show that both the volume and volatility of equity purchases by U.S. investors increased substantially after these countries opened their stock markets to foreign investors (Table 10).

Table 9.

Total Returns on Equity in Selected Emerging Market Countries: Before and After Opening to Foreign Investors1

(In percent at annual rate)

article image
Source: IFC Emerging Markets data base.

Time period is 1976, quarter I to 1994, quarter II. The actual dates of opening to foreign investors are as follows: Greece—December 1988; Portugal—December 1988; Turkey—August 1989; Jordan—December 1988; Argentina—October 1991; Brazil—May 1991; Chile—December 1988; Colombia—February 1991; Mexico—May 1989; Venezuela—January 1990; India—November 1992; Korea—January 1992; Malaysia—December 1988; Pakistan—February 1991; Philippines—October 1989; Taiwan Province of China—January 1991; Thailand—December 1988. For analytical purposes, and to capture the fluctuations in the market caused by changes in investor expectations, the opening date was taken to be four months prior to the actual date.

Volatility is measured by the standard deviation of the percent change in the IFC total return indices.

Table 10.

U.S. Net Purchases of Foreign Equity in Selected Countries: Before and After Opening to Foreign Investors

(In millions of U.S. dollars)

article image
Source: U.S. Treasury, Treasury Bulletin.

Measured by the standard deviation.

11

Includes reported private placements and notes issued under the Euro-medium-term note programs. The figures differ from OECD estimates, which have a narrower coverage.

12

Throughout this chapter, yield spreads refer to the difference between the yield on a bond at the time of issuance and the yield on U.S. Treasury securities of comparable maturity or other comparable government securities if the bond is issued in other currencies. The U.S. Treasury security and other comparable government securities are used as a proxy for a risk-free return.

13

These estimates can vary somewhat depending upon whether bondholders decide to take early redemption options for some bonds.

14

Global bonds are issued simultaneously in several major international markets and allow issuers to tap into broader demand and obtain lower rates than those available in a single market. Some market participants estimated that Argentina was able to reduce the interest rate on the funds raised through the global issue by as much as 30 basis points.

15

The Yankee bond market is the domestic U.S. market for U.S. dollar-denominated bonds issued by nonresident entities. Yankee issues are subject to SEC registration and disclosure requirements.

16

Simple auto-regression tests of market efficiency were run on the prices of U.S. dollar-denominated sovereign bonds of Argentina, Brazil, Hungary, Mexico, Nigeria, the Philippines, Turkey, and Venezuela. In all cases, these tests suggest that the markets are inefficient.

17

In both countries, there was substantial deterioration in economic conditions and some political instability over the period tested.

18

Since monthly price indices for subinvestment-grade U.S. corporates comprise only bonds with seven-year maturities, the analysis was conducted using Brady par bonds for the countries that have issued these securities. These bonds have remaining maturities of 25 years or more. A relatively constant differential between the prices of Brady bonds and U.S. corporate securities would largely represent a yield curve effect arising from the difference in the maturities of these two sets of bonds. For the Eurobonds analyzed, an adjustment had to be made to the prices of the developing sovereign bonds to factor in the effect of a declining yield curve as these securities moved closer to maturity.

19

Unpublished work by William Cline at the Institute for International Economics has found on the basis of pooled cross-sectional data that countries that previously restructured bank debt and those with higher inflation and lower export or per capita GDP growth tend to have higher spreads on their new bonds. Preliminary regression analysis by IMF staff relating movements over time of country bond spreads to various indicators of economic fundamentals, however, did not produce significant results, in part owing to the limited time period for which data are available.

20

A notable exception is the Venezuelan Eurobond. The lack of response in the yield spread for this bond to the deterioration in the country’s recent economic performance may reflect the fact that the issuer of the bond is the state-owned oil company, which may be considered a better credit risk because of its external assets.

21

An ADR is a U.S. dollar-denominated equity-based instrument backed by shares in a foreign company held in trust. ADRs are traded like the underlying shares of stock on major U.S. exchanges or in the over-the-counter market.

22

A GDR is similar to an ADR, but it is issued and traded internationally.

23

Emerging market mutual funds’ investment in developing country bonds has been limited. In 1993, net assets of fixed income funds amounted to only $8.5 billion. These figures are based on information provided by Emerging Market Funds Research, Inc., and Lipper Analytical Services, Inc. Since funds that have invested less than 60 percent of their portfolio in emerging markets are not included, developing country assets purchased by mutual funds may be understated. On the other hand, to the extent that emerging market mutual funds usually hold part of their assets in cash or developing country assets, the net asset value of these funds may overstate actual investment in emerging markets securities.

24

Net equity purchases are estimated by deflating changes in the net assets of each regional fund by the corresponding IFC investable share price index. The estimates are subject to a wide margin of error, especially because the country weights used for the IFC’s regional and global indices may differ from the country composition of the equities held by the funds.

25

As the total for 1992 was affected by a large credit to Saudi Arabia, the underlying growth was probably higher. The figure excludes loans guaranteed by export credit agencies.

27

Consider an example of two very risky assets with the same average return, but that always move in the opposite direction, that is, their returns are perfectly negatively correlated. A portfolio divided equally between these two assets would in principle involve no risk, as the movements in the two asset returns would always offset each other. Thus, a portfolio composed of both assets would yield the same expected return, while involving less risk than a portfolio consisting entirely of either one of the assets. These points have been applied to portfolios with emerging markets assets in a number of research papers, such as Campbell (1993).

28

This view of the structure of a financial market is reviewed in Grossman and Stiglitz (1980).

29

This approach is developed in Kyle (1985) and Campbell and Kyle (1993).

30

The ability of large traders to conceal their trades would be expected to diminish over time as the number of large traders increased.

31

The indices of equity prices were drawn from the Emerging Markets data base of the International Finance Corporation.

32

U.S. Department of Treasury, Treasury Bulletin.

  • View in gallery

    Private Market Financing to Developing Countries

    (In millions of U.S. dollars)

  • View in gallery

    Yield Spreads at Launch for Unenhanced Bond Issues by Developing Countries1

    (In basis points)

  • View in gallery

    Secondary Market Yield Spreads on U.S. Dollar Denominated Bonds by Selected Developing Countries

    (In basis points)

  • View in gallery

    Comparison of Yields of Sovereign Bonds with Yields on U.S. Corporate Bonds

    (In percentage points)

  • View in gallery

    Comparison of Sovereign Bond Spreads

    (In basis points)

  • View in gallery

    Comparison of Movements in Spreads and Economic Variables for Mexico and Venezuela

  • View in gallery

    Comparison of Movements in Spreads and Economic Variables for Hungary and Turkey

  • View in gallery

    Emerging Market Mutual Funds

    (In billions of U.S. dollars)

  • View in gallery

    Share Price Indices for Selected Markets In Latin America

    (IFC Weekly Investable Price Indices, December 1988 = 100; in U.S. dollars)

  • View in gallery

    Share Price Indices for Selected Markets In Asia

    (IFC Weekly Investable Price Indices, December 1988 = 100; in U.S. dollars)

  • View in gallery

    Weekly Volatility of Total Returns on Bonds and Equities for Selected Countries1

    (In percent)

  • Campbell, H., “Portfolio Enhancement Using Emerging Markets as Conditioning Information,” paper presented at World Bank Symposium on Portfolio Investment in Developing Countries, Washington, D.C., September 1993.

    • Search Google Scholar
    • Export Citation
  • Campbell, J.Y., and A.S. Kyle,Smart Money, Noise Trading, and Stock Price Behavior,Review of Economic Studies, Vol. 60 (January 1993), pp. 134.

    • Search Google Scholar
    • Export Citation
  • Cardoso, E., and R. Dornbusch,Foreign Private Capital Flows” in Handbook of Development Economics, Vol. 2, ed. by Hollis Chenery and T.N. Srinivasan (Amsterdam; New York: North-Holland, 1988).

    • Search Google Scholar
    • Export Citation
  • Claessens, S,Alternative Forms of External Finance: A Survey,World Bank Research Observer, Vol. 8 (January 1993), pp. 91117.

  • Claessens, S, M. Dooley, and A. Warner,Portfolio Capital Flows: Hot or Cool?World Bank Discussion Paper, August 27, 1993.

  • Collyns, Charles, and others, Private Market Financing for Developing Countries (Washington: International Monetary Fund, 1993, 1992, and 1991).

    • Search Google Scholar
    • Export Citation
  • Dooley, Michael,An Analysis of the Debt Crisis,IMF Working Paper, WP/86/14 (Washington: International Monetary Fund, 1986).

  • Edwards, Sebastian,Capital Flows, Foreign Direct Investment, and Debt-Equity Swaps in Developing Countries,National Bureau of Economic Research Working Paper, No. 3497 (October 1990).

    • Search Google Scholar
    • Export Citation
  • Goldsbrough, David J.,Investment Trends and Prospects: The Link with Bank Lending” in Investing in Development: New Roles for Private Capital, ed. by Theodore T. Moran and others (New Brunswick, Connecticut: Transaction Books, 1986)

    • Search Google Scholar
    • Export Citation
  • Goldstein, Morris, and M. Mussa,The Integration of World Capital Markets,IMF Working Paper, WP/93/95 (Washington: International Monetary Fund, 1993).

    • Search Google Scholar
    • Export Citation
  • Goldstein, Morris, and others, International Capital Markets Part II, Systemic Issues in International Finance (Washington: International Monetary Fund, August 1993).

    • Search Google Scholar
    • Export Citation
  • Goldstein, Morris, and others, International Capital Markets: Developments, Prospects, and Key Policy Issues (Washington: International Monetary Fund, 1994).

    • Search Google Scholar
    • Export Citation
  • Graham, Edward M., and Paul R. Krugman, “The Surge in Foreign Direct Investment in the 1980s” in Foreign Direct Investment, ed. by Kenneth Froot (Chicago: University of Chicago Press, 1993).

    • Search Google Scholar
    • Export Citation
  • Grossman, S.J., and J.E. Stiglitz,On the Impossibility of Informationally Efficient Markets,American Economic Review, Vol. 70 (1980).

    • Search Google Scholar
    • Export Citation
  • International Finance Corporation (1994a), Emerging Markets Factbook (Washington: IFC, 1994).

  • International Finance Corporation (1994b), Quarterly Review of Emerging Markets, (First Quarter 1994).

  • International Organization of Securities Commissions (IOSCO), Report of the Development Committee on Internationalization (Montreal: IOSCO, October 1993).

    • Search Google Scholar
    • Export Citation
  • Kuhn, Michael G., and others, Official Financing for Developing Countries (Washington: International Monetary Fund, April 1994).

    • Sea