Percentage of GDP

In most developing countries the financial system is dominated by banks, and nonbank intermediation is relatively underdeveloped. There are several reasons why a developed bond market is useful in mobilizing long-term funds for investment: • Systemic risk. One of the most important (and least appreciated) benefits of bond markets is the dispersion of risk. The absence of active and open securities markets in developing countries places excessive risk on bank-based financial systems (McKinnon 1988). This dispersion reduces systemic risk and thus enhances the ability of the financial system to cope with shocks. • Maturity. Because of the dispersion of risks among savers, it is easier to attract long-term finance through bond markets than from commercial banks. Developed bond markets provide investors with the long-term funds required to build infrastructure and housing, and savers with the opportunity to purchase high-yield financial instruments (Gelb and Honohan 1989). Banks in developing countries have limited ability to mobilize long-term resources and to finance long-term investment. The development of securities markets helps reduce the mismatch between the short-term deposits of the banking system and evergreen credit lines, thereby enhancing the soundness of the financial system (Agtmael 1984). • Liquidity. Because bonds can be bought or sold in the secondary market, they have a lower liquidity premium than commercial bank debt, thus reducing the cost of capital. However, high marketability requires a large trading volume and a large number of dealers. These conditions are not met in many developing country bond markets. • Mobilization of savings and innovation. There have been unparalleled innovations in bond markets in recent years, such as mortgage and asset-backed securitizations, stripping, and derivatives, which have greatly increased the flexibility of financial instruments. These structures tap into new market segments of issuers and purchasers alike and therefore enhance the mobilization of funds. Overall, innovation can increase the amount of savings channeled into productive investments. • Allocation of resources. The bond market does an effective job of selecting and monitoring productive investments. The continuing assessment and monitoring of risks conducted by impartial and informed credit rating agencies distinguishes the bond market from the loan market and helps to foster good corporate conduct and governance. Thus bond markets can make the financial and economic system more competitive and efficient by lowering the initial cost of capital (Agtmeal 1984). • Policy instruments. The availability of well-developed bond markets provides a useful source of funds to finance government deficits (which may reduce the government’s reliance on money financing) and offers the instruments to conduct effective monetary policy (see box 1A.2). • Competition and efficiency in financial services. Evidence suggests that active primary and secondary markets for bonds and equities, by providing competition for the banking sector, can improve the performance of the financial system. According to an IMF (1997) study of innovations in OECD capital markets, intermediation costs have been sharply reduced by the substitution of direct transactions in securities for bank credits, by reduced commissions, and by increased competition.

Access by developing country governments to international bond markets has increased markedly in recent years. The government debt of 51 developing countries is now rated by the main international credit agencies (just 10 were rated in 1990), and 25 developing countries achieved investment grade status on their sovereign debt (prior to the East Asian crisis). Publicly guaranteed debt from private sources in these countries ranges from 0.5 percent of GDP in Malta to 50.7 percent in Panama, with an average of 6 percent. Many governments in developing countries are now able to choose between borrowing in local currency in domestic markets and borrowing in foreign currency in international markets. Some governments have issued debt linked to foreign currencies in domestic markets, such as the Mexican tesobonos (Guidotti and Kumar 1991). In deciding where to borrow, governments in developing countries typically face a highly segmented market, with policy and institutional impediments to capital flows, asymmetric information, and differing levels of risk all contributing to the segmentation between domestic and foreign markets. Governments that have access to international private sources of finance in effect face two distinct supply schedules for funds. Both the domestic and foreign supply schedules are likely to be upward sloping, reflecting (in both domestic and international markets) rising risk perceptions with increased borrowing and (in domestic markets) a limited supply of savings. (See, for example, Eaton 1989; Cohen 1991; Eaton, Gersovitz, and Stiglitz 1986; Tanzi and Blejer 1988; Stiglitz 1988; and Calvo and Guidotti 1993.) There is, therefore, a rationale for borrowing to be distributed between domestic and foreign markets so as to equalize marginal cost. However, in deciding whether to borrow abroad or at home, governments (which are typically by far the largest borrower) go beyond direct cost and maturity comparisons to consider the implications for the economy. The government’s choice can affect the cost of foreign borrowing for the economy as a whole, the country’s creditworthiness, the foreign currency exposure of the government and the private sector, the development of the domestic capital market, and the conduct of monetary policy: • Cost of borrowing. In most developing countries the ex ante return to capital is likely to be higher than the cost of foreign borrowing, reflecting both capital scarcity and high volatility in developing markets. However, international lenders may lack information on domestic private sector borrowers, and there may be a lack of creditworthy companies able to borrow on the required scale. In this case it is cheaper at the margin for the government to borrow rather than for individual private firms to do so, reflecting the government’s greater access to the market. At the same time, private sector borrowers can rely on domestic banks, which generally face lower costs than do international lenders in evaluating the riskiness of domestic firms. By borrowing abroad the government avoids crowding out firms on the domestic market and reduces the cost of capital for the whole economy. Of course, whether this turns out to be a net benefit depends on whether the use of funds by the public sector is as efficient as in the private sector. Also, government borrowing can improve the familiarity of international lenders with the country and establish a benchmark that can be used to price loans to private companies, so that initial borrowing by the government may reduce the ultimate cost of borrowing by the private sector. • Foreign debts and creditworthiness. Defaults by private sector borrowers can affect a country’s credit standing in the eyes of foreign investors. If relatively risky borrowers can gain access to international capital markets (for example, during a period of unusually high liquidity when investors may accept higher levels of risk to maintain yields), the government may decide to borrow more abroad rather than from domestic markets, thus lessening pressures on interest rates and increasing incentives for firms to borrow domestically. • Currency exposure. Since foreign borrowing is usually in foreign currency it implies an increase in foreign exchange exposure (Sachs, Tornell, and Velasco 1996). (This is not always true, however. For example, domestic bond markets in Brazil and Indonesia are accessible by foreigners, and Argentina and South Africa have issued local currency bonds.) However, where public enterprises are major exporters or where tax revenues from the private sector are significantly affected by shifts in the exchange rate (in economies reliant on primary commodities), borrowing in foreign currency may reduce the country’s net exposure to currency risk. For example, a government that is dependent on dollar-denominated exports for its revenues can borrow abroad in dollars. Any losses (or gains) in export receipts from currency fluctuations would then be in part balanced by gains (or losses) in the cost of debt service payments. • Institutional development. Government participation may be the surest and fastest way of developing the domestic bond market, thus opening up new and more versatile financing instruments for domestic companies and providing a new channel to reward and retain domestic savings. • Monetary policy. The government may also decide to shift the mix of its borrowing between domestic and foreign sources in the short term depending on its stabilization objectives, relying more on domestic borrowing when the economy is overheating and more on foreign borrowing when domestic activity slackens. (On the theoretical aspects of the mix between foreign and domestic bonds, see Bohn 1990. On macroeconomic management aspects, see Anand and van Wijnbergen 1988 and Cuddington 1997) The possible advantages enumerated above of governments borrowing abroad instead of at home (lower cost, longer maturities, avoiding crowding out of private firms, better management of currency exposures, and stabilization) will tend to diminish as capital markets become more integrated, incomes and capital intensity rise, economic volatility diminishes with increased diversification of the economy, and macroeconomic management improves. These developments will also tend to equate the foreign and domestic cost of funds and result in a more efficient allocation of capital internationally.