Appendix Financial Systems and Growth
Total Page:16
File Type:pdf, Size:1020Kb
Appendix Financial Systems and Growth he following reviews the key channels through ther positive or negative, the reduced risk of illiq- T which the financial system is related to eco- uidity enhances productivity. By encouraging in- nomic growth. This appendix also reviews the ques- vestment in longer-term projects with higher re- tions of whether there is empirical evidence concern- turns, which in the absence of capital markets ing the relationship between financial sector would not be undertaken, the average return on in- development and economic growth and, if so, vestment in the economy increases and with it so whether there is any evidence for a causal relation- does productivity. ship. Finally, this section looks at whether a particu- The financial system facilitates the acquisition of lar structure of the financial system is superior in information and the allocation of resources. High in- terms of growth prospects. formation costs can hinder the flow of capital to its most productive use. The financial system can lower the costs of obtaining information for an individual The Role of Financial Intermediaries in two ways. By grouping together, i.e., forming in- and Capital Markets termediaries, individuals can spread any fixed costs of obtaining information, and in well-developed sec- The primary role of the financial system is to fa- ondary markets published prices are an important cilitate transactions and the allocation of resources source of information. across agents and over time. The role for financial Capital markets provide a way for agents to moni- intermediaries arises in the presence of market im- tor managers in the firms in which the agents have perfections, such as information asymmetries, invested; this can increase the level of investment which give rise to transaction costs and the cost of and also the productivity of investment. Individual obtaining information. By reducing these costs, investors need a way to ensure that the funds they more transactions can occur, improving the alloca- have invested are being properly managed; however, tion of resources. Levine (1997) discusses the basic such monitoring can be very costly for the individual functions of financial intermediaries as well as cap- investor. Well-developed collateral and financial ital markets and the channels through which they contracts can lower monitoring and enforcement enhance economic growth. Each function can in- costs. Additionally, financial intermediaries can fluence growth through its impact on investment spread the monitoring cost across groups of individ- decisions either by increasing the level of invest- uals. Bencivenga and Smith (1991) have shown that ment or by increasing the efficiency of those financial arrangements that improve corporate con- investments and thus the rate of technological trol lead to higher levels of investment and faster improvement. growth. One basic function of capital markets is to facili- Another role of the financial system is to mobi- tate the trading, hedging, diversifying, and pooling lize savings; by pooling the savings of individuals, of risk. An important risk faced by agents consider- firms can undertake large investment projects not ing an investment project is that they will not be possible without the resources of several investors. able to sell their assets to meet current consumption In the presence of economies-of-scale, the produc- needs. By increasing the ease and speed with which tive capacity of the economy is increased. Finally, assets can be converted into purchasing power, cap- financial intermediaries facilitate the exchange of ital markets increase liquidity and reduce the risk goods and services by reducing transaction costs. of illiquidity for an individual investor. While the The greater the ease with which goods and services impact of increased liquidity on the overall level can be traded, the higher the degree of specializa- of savings, and therefore on investment, may be ei- tion, which in turn implies higher productivity. 32 ©International Monetary Fund. Not for Redistribution Appendix Empirical Evidence on Financial Figure A.1. Financial Development and Market Development and Growth Real Per Capita GDP in Country Groups, There is a wealth of empirical research investi- 1985 gating the relationship between economic growth and financial system development. Much of the re- search has taken place during the 1990s and was spurred by the influential studies of King and Levine (1993a) and (1993b). Those studies exam- ined the relationship between economic growth and measures of financial sector development using the familiar methodology of cross-country growth studies.49 Figure A.1 shows the variation of specific indica- tors of financial development in 1985 across 116 countries divided into four groups by real GDP per capita. The first is a measure of the size of the fi- nancial intermediation sector; the second measures the extent to which commercial banks rather than central banks are allocating credit; while the third and fourth indicators concern the allocation of Source: Levine (1997). credit, measuring the portion of credit allocated to 1Depth = Ratio of liquid liabilities of the financial system to GDP. the private sector and the amount of credit extended 2Bank = Ratio of bank credit (domestic deposit money banks) to combined bank credit and central bank credit. relative to GDP. All these measures have a strong 3Private = Ratio of claims on the nonfinancial private sector to positive relationship to GDP and show that finan- total domestic credit. 4Privy = Ratio of gross claims on private sector to GDP. cial markets are larger and more market oriented in 5RGDP = Real GDP per capita in U.S. dollars. richer countries than in poorer countries, with a much smaller proportion of credit being extended by central banks and a much greater proportion being extended to the private sector in richer coun- tries. Regression analysis showed that these indica- longer samples of adequate data, Vector Autore- tors have a strong positive and economically im- gression (VAR) techniques may be employed, as in portant relationship with average real GDP growth, Wachtel and Rousseau (1995) and Wachtel (2001), the rate of capital accumulation, and total produc- and these studies find evidence of a causal relation- tivity growth. ship. A causal relationship is also found using Moreover, the studies show that the initial value panel VAR techniques in cross-country studies, of financial depth proves to be a significant predic- see for example Beck, Levine, and Laoyza (2000). tor of future growth, capital accumulation, and pro- The consensus that emerges from these studies is ductivity growth. This result is important because it that there is a robust causal relationship between fi- addresses the issue of causality and suggests that nancial development and economic growth. As the development of the financial system can in- noted by Khan and Senhadji (2000), however, the crease economic growth rates. The evidence on size of the measured effect varies with different in- causality, however, is not conclusive because there dicators of financial development, estimation are a number of econometric problems associated method, data frequency, and the functional form of with the methodology, one of which is the issue of the relationship.50 simultaneity bias. A number of studies have ap- plied various approaches to deal with this issue, and, broadly speaking, the conclusions are robust to the methodology employed. For countries with 50Much of the work on this topic assumes a linear relationship between financial market development and economic growth. One exception is Khan and Senhadji (2000) who allow for a sim- ple form of nonlinearity by specifying a quadratic relationship. A 49The standard regression model with the growth of per capita statistically significant negative coefficient is found on the qua- real GDP as the dependent variable is estimated using panel data. dratic term, which could suggest that there exists an optimum The explanatory variables include conditioning variables, such as level of financial development. Another possibility is that there the log of initial real per capita GDP and initial secondary school may exist a threshold effect—that countries need to reach a cer- enrollment rate, among others, and a measure of financial sector tain level of financial development before there are effects on development (the variable of interest). growth, as argued by Berthelemy and Varoudakis (1996). 33 ©International Monetary Fund. Not for Redistribution APPENDIX Characteristics of Financial Markets absence of a bond rating, there is an increased re- and Growth liance on internal sources of financing. This result would be consistent with government measures to Thus far, the discussion has focused on evidence increase transparency. concerning the relationship between broad aggre- Finally, the question of whether the general struc- gates that measure various aspects of financial devel- ture of the financial system has implications for opment and growth. Results from such studies are economic growth has received considerable atten- quiet, however, on specific actions that policymakers tion, as reflected in studies such as the one by might undertake in order to promote the develop- Claessens, Djankov, and Klingebiel (2001) and by ment of the financial system. Nevertheless, there are Demirguc-Kunt and Levine (1996). These studies a number of studies that examine the relationship be-