1
UNIVERSITY OF GRONINGEN
Faculty of Economics and Business
MSc Business Administration, Specialization Finance
STOCK MARKETS, BANKS AND ECONOMIC
GROWTH
Supervising professor:
Prof. Dr. Kasper Roszbach
2
Abstract
This paper investigates the empirical relationship between financial structure, financial development and economic growth for the sample of 48 countries through two sample periods, 1980-1995 and 1996-2009. The paper provides an extended analysis for the research paper of Levine (2002); it re-examines the paper of Levine (2002) for the initial sample period of 1980-1995 and extends the analysis for the period 1996-2009, using the simple and the full conditioning information set as sets of control variables. I find evidence of a significant long-term relationship between the level of financial development and the real GDP growth for both investigated periods. I do not find evidence of the robust link between financial structure and economic performance. None of the variables used as a robustness check for financial structure - the stage of financial development and the legal and the contractual environment indicates significant impact on economic growth. The results for both sample periods are consistent with the finance view; they do not support the bank-based, the market-based and the law view. This conclusion corresponds with the results of Levine (2002).
4 1 Introduction
The relation between the financial development and the economic growth has become one of the most extensively investigated subjects in the finance literature. Since burgeoning research papers of Schumpeter (1912), Gurley and Shaw (1955) as well as Goldsmith (1969) provide many current academici, e.g. Holstrom and Tirole (1993) and Sirri and Tufano (1995), evidence that the finance-growth nexus cannot be rejected. On the contrary to these, Robinson (1952), Meier and Seers (1984) and Lucas (1988) etc. disagree about the role of financial sector in promoting economic growth.
A key distinguishing feature of financial markets compared to banks lies in the ability to form prices, so called “information feedback”. This function is based on the ability of markets to provide the information to companies (the supply side function) and the ability of companies to reflect this information in their decision making process (the demand side function). The effectiveness of the supply side of the information feedback is reduced by the weak legal environment and the poor contract enforcement. Under these circumstances, according to Rajan en Zingales (1998), banks prove a better position since they have bigger enforcement power than financial markets. In the poor legal and contractual environment is the “information feedback” of the markets lower, so is the value of the information affecting the real decision making process of firms. Boot and Thakor (1997) stress the efficiency of banks in the economies where the majority of firms are affected by the moral hazard problem.
5 the significant impact of financial structure, and the law system on the level of economic performance. His results are therefore consistent with the finance view and not with the bank-based view, the market-based or the law view. In order to re-estimate and afterwards extend the analysis, the research question can be formulated as follows:
Does overall financial development support economic growth? Does financial structure matter in promoting economic growth? If so, is the financial architecture anchored on markets more beneficial
in promoting economic performance than the one centered on banks? Does the legal and the contractual environment, most of the time underestimated components of overall financial development, have any influence on economic growth?
In order to answer this question, in line with the methodology of the Levine (2002) I provide a cross-sectional analysis on 48 countries over the periods 1980-1995 and 1996-2009. Due to the fact, that no universally accepted formulation of the bank-based and the market-based system is defined, four measures of financial structure are necessary to be constructed; namely STRUCTURE-ACTIVITY, STRUCTURE-SIZE, STRUCTURE-EFFICIENCY and STRUCTURE-AGGREGATE. In order to prevent the anomalies while applying the measures of financial structure, I use a dummy variable in the regression to identify highly underdeveloped countries. Furthermore, there are four proxies which are used as measures of financial development, so that the level of financial services; FINANCE-ACTIVITY, FINANCE-SIZE, FINANCE-EFFICIENCY, FINANCE-AGGREGATE. In order to obtain the requested relationships, Ordinary least squared method (OLS onwards) method is applied to these variables. Furthermore, there are several control variables included in the regression which are grouped into the simple conditioning information set and extended, full conditioning information set.
6 This paper contributes to the research literature on the financial structure-growth nexus in two ways. Firstly, besides re-estimating the results of Levine (2002), the paper extends the analysis applied for the period 1980-1995 for the following period of years, 1996-2009 and simultaneously, it enables to provide a wide range of comparisons of the results both among the sample periods and cross-sectionally. Secondly, in line with the analysis for the years 1980-1995, the paper applies the conglomerate variables of financial structure in the analysis for the years 1996-2009 as well and therefore brings further evidence on suitability of these variables for the usage in the analysis.
The remainder of this paper is organised as follows. Section 2 reviews previous theoretical and empirical studies on this topic. Section 3 contains variable definitions and the data source. Section 4 presents to be tested hypotheses. In section 5 methodology is provided. In section 6, I present the results. Finally, section 7 concludes.
2 Literature review
2.1 Theoretical considerations
Financial markets and banks fulfill vital functions in the economy; they are important producers of the information, enable to monitor firms in an efficient way, are important facilitators of risk sharing and most importantly, they enable to accumulate and form the capital. In particular, a key attribute of financial markets as a feature that distinguishes them from banks is that they have an ability to form equilibrium prices which provide valuable information about the prospect of investment opportunities to real decisions of firms.
7 Another function of banks is presented by Greenwood and Jovanovic (1990), Bencivenga and Smith (1991), de la Fuente and Marin (1996) and Allen and Gale (1999) who argue that by pooling the cross-sectional, inter-temporal and liquidity risk banks are able to provide financing and monitor risky investments, therefore contribute to investment efficiency and promote economic growth. Stulz (2000) is strongly persuaded about an important role of banks in the innovative projects with the stage financing. Banks are more effective sources of external financing to innovative investments with the stage financing since they are more open to provide additional funding to firms as investment projects develop and enable investors to be informed about the run of the project.
Finally, Sirri and Tufano (1995) claim that banks play an important role in the process of exploiting the economies of scale and the scale of industrial growth by mobilizing capital, but only under the condition that bank activities are not affected by regulatory restrictions. Even though Chakraborty and Ray (2006) in their teoretical study do not confirm that the bank-based system has positive effect on growth, they conclude that the bank-based system outperforms the market-based system in many dimensions. They state that in the bank-based economies per capita income is higher and the income inequality lower. Growth of the manufacturing sector and better wealth distribution enable to support the flow of the investments into the manufacturing sector more intensively since banks provide external finance to more entrepreneurs.
labour-8 intensive sectors than to the innovative, strategic and R&D-based sectors. Financial markets assure this problem is avoided.
Opposed to the proponents of the bank-based system, there are many analysts who tout the advantages of the market-based view. Whereas Holstrom and Tirole (1993) highlight a significant role of well-functioning, big and liquid markets in fostering incentives to research firms, Levine (1991) and Obstfeld (1994) see their great importance in facilitating the risk management of firms. Furthermore, Jensen and Murphy (1990) stress a non-negligible function of markets in enhancing corporate governance by enabling to derive the manager compensation from the firm performance and by easing the takeovers. Hellwig (1998) mentions the possibility that the monitoring bank and the monitored firm will collude and Hellwig (1991) stresses the fact, that the moral hazard problem may appear on the side of the monitoring bank.
Levine (1991) and Bencivenga and Smith (1991) say that liquid financial markets support investment incentives to finance long-term projects since atomic markets provide them the flexibility to sell their stakes as a reaction to the new information so that e.g. in case they have a need for their savings before the end of the project, they can sell a stake in a project or in a firm easily. Furthermore, Bhide (1993) concludes that investors in liquid atomic markets have lower incentives to exercise corporate governance control in projects or companies in which they have stake in since these investors have a possibility to sell their stake easily. Stiglitz (1985) states that lower incentives of investors to monitor firms lie in the ability of effective financial markets to reveal the information in the stock prices promptly. In line with the conclusions of Boot (1993), banks are not facing this problem due to their nature of their existence.
There are scholars who do not find a debate about the competitive position of banks and financial markets relevant. They rather stress the fact that effective cohesion and cooperation between these financial arrangements can contribute to economic growth. Chakraborty and Ray (2006) provide theoretical evidence that the growth rate depends on the efficiency of legal and financial institutions and not on the specialization of the economy that they work in.
9 mechanisms that govern debt and equity transactions are definitely more useful way to distinguish financial systems than distinguishing between the bank-based and the market-based systems. La Porta et al. (1998) examines presence of legal rules covering the protection of legal rights of shareholders and creditors in the sample of 49 countries, aggregating them into indices of creditors` and shareholders` rights and proposes measures of the enforcement quality. Coffee (2000) and Johnson et al. (2000) provide further investigation into the topic of legal rules in the legal families and the investor protection.
2.2 Empirical considerations
Levine and Zervos (1998) using data on 47 countries from 1976 through 1993 empirically investigate whether measures of stock market liquidity, size, volatility, and integration with world capital markets are robustly related to the current and future rates of economic growth. They provide evidence that initial levels of both stock market liquidity, measured by the value of stock trading relative to the size of the market, and banking development, measured by bank loans to private enterprises divided by GDP are significantly correlated with the current and future rates of economic growth. Furthermore, they conclude a strong, positive link between the initial level of financial development and the current level of economic growth.
Roussel and Wachtel (2000), in their paper provide analysis on 47 countries applying the GMM (Generalized Method of Moments) method over the panel data vector autoregression. They state that the activity of stock market, measured by the value traded, rather than the size of equity market, measured by the market capitalization, has a leading role in promoting economic development. Furthermore, the relationship between banks and economic growth, represented by the variable total liquid liabilities of financial intermediaries (M3) is positive.
10 Furthermore, King and Levine (1993) using data from 80 countries in the cross-country analysis conclude that the level of bank development has a positive effect on economic growth. However, since the analysis focuses primarily on banks as representants of financial system and equity markets are not considered in it, it is discussable whether the positive relationship between bank development and the level of economic growth will stay robust when controlling for the presence of equity markets as well.
In order to examine the relationship between banks, stock markets and economic growth Beck and Levine (2004) apply the methodology of Levine and Zervos (1998), however, in order to avoid statistical weaknesses of their method, they use GMM (Generalized Method of Moments) for dynamic panel data, controlling for the omitted variable bias and the simultaneity bias. Sample period covers the period of 1976-1998. Although they initially find out that both banks and stock markets independently have a positive effect on the level of economic success, after applying several estimation techniques and control variables they conclude that overall financial development matters for the economic growth but are unable to provide evidence about which institution spurs economic growth with higher intensity.
Tadesse (2000), using industry-level data from a panel of 36 countries over the period 1980 through 1995, provides evidence that development of financial services plays a key role in the position of banks and stock markets on the real economic performance. Similarly as Boyd and Smith (1996, 1998), but applying their analysis to the industries, Tadesse (2000) states that whereas industries supported by bank-based financial systems grow faster across countries with less developed financial systems, in countries with developed financial sectors market-based systems boost economic growth more intensively. However, the second conclusion is applicable only for countries in which big firms are present as a major group among firms. Moreover, in line with Rajan and Zingales (1999) Tadesse (2000) claims that in the environment with weak legal and institutional preconditions bank-based systems have a comparative advantage. Hence, market-based systems fare well in stronger contractual and legal environments.
3 Variable definitions
In the regression conglomerate measures of financial structure and financial development are used as explanatory variables and economic growth as explained variable. The usage of these variables is in consistence with Levine (2002).
11 of the economic growth, e.g. Levine and Zervos (1998), Rousseau and Wachtel (2000); Arestis et al. (2001), Beck and Levine (2004).
Financial structure indicates whether an economy is bank-based system or the market-based. The
advantage of providing such a cross-country analysis on a rich research sample lies in obtaining international comparisons. However, this broad analysis does not enable to include some indicators that can potentially have an influence on the final results. Therefore, I construct aggregate indicators of financial structure expressed by relative measures which incorporate size, activity, efficiency of bank and financial markets on bank activities.
STRUCTURE-SIZE expresses the relation between the size of stock markets, represented by the
logarithm of the market capitalization ratio and size of banks, represented by the logarithm of the bank credit ratio. Market capitalization ratio captures the value of the shares of domestic firms listed on the domestic exchanges as a share of GDP. Bank credit ratio equals the share of deposit money bank claims on the private sector, including credits to deposit taking institutions according to the clasification provided by the International Monetary Fund, excluding credits to governments and public enterprises.
The relation between the activity of stock markets, represented by the logarithm of the total value traded ratio and activity of banks, interpreted by the logarithm of the bank credit ratio is expressed by the STRUCTURE-ACTIVITY measure. Total value traded ratio equals the value of shares of domestic firms traded on domestic exchange divided by GDP, this measure is used to capture the liquidity of the stock market. Bank credit ratio, similarly, as in the Structure-size measure, expresses the share of claims of deposit money banks to the private sector, including credits to deposit taking institutions and excluding credits to governments and public enterprizes.
The STRUCTURE-EFFICIENCY indicator equals the efficiency of stock markets relative to the efficiency of banks. The efficiency of stock markets is captured by firstly, their liquidity, the total value traded ratio, and furthermore by the turnover ratio, which equals the value of shares traded on domestic stock market to the value of shares of domestic companies listed on stock market. As a measure of efficiency of banks, the measure of overhead costs is used, expressed by the ratio of overhead costs in the banking system to the total assets in the banking sector.
12 In conducting his research, Levine (2002) provides the analysis on the variable of STRUCTURE-REGULATORY as well. 1 Due to the limited access to the data and therefore limited comparability of the data during the sample research periods, analysis on the STRUCTURE-REGULATORY variable is not included in this paper.
The financial development measures
The measures of overall financial sector development capture the level of development of financial services in countries and are main evaluating measures of the financial services view. Here, the discussion about the importance of bank and market system is not relevant.
The FINANCE-SIZE indicator is calculated by multiplying logarithm forms of the market capitalization ratio, an indicator of size of the domestic stock market and private credit ratio as a measure of size of financial intermediaries. Even though e.g. Levine and Zervos (1998) conclude that the market activity rather than the market size has a positive effect on the level of economic growth, in the analysis market capitalization ratio rather than total value traded ratio is applied in order to retain the consistency.
The FINANCE-ACTIVITY indicator is obtained by multiplying the logarithms of activity of financial markets, represented by total value traded ratio and the activity of banks, represented by private credit ratio. In comparison with the structure-activity measure where bank credit ratio is used to capture the activity of banks, in order to cover the activity of all financial intermediaries credits provided by non-deposit money banks are also necessary to be included into the index. However, similarly as in case of bank credit ratio, credits to the public sector are excluded.
The FINANCE-EFFICIENCY indicator measures efficiency of the financial sector. It is calculated as the ratio of total value traded ratio, which represents the efficiency of financial markets, and overhead costs as a measure of efficiency of banks, expressed by the ratio of overhead costs in the banking system to the total assets in the banking sector.
Similarly, as in case of STRUCTURE-AGGREGATE, the FINANCE-AGGREGATE index is constructed from the first principal components of the financial development variables
FINANCE-ACTIVITY, FINANCE-SIZE and FINANCE-EFFICIENCY.
1 The variable measures a degree to which banks operate in securities, insurance and real estate activities and
13
The simple conditioning information set includes initial real per capita GDP, in this paper 1980 and
1996, initial level of the number of years of schooling in the working age population. These variables are inevitable to be captured in the analysis because of importance of human capital in the promotion of growth.
In order to include macroeconomic indicators and indicators of political situation in a country, the full
conditioning set is applied in the analysis. Besides the variables of simple conditioning set it contains
the variables of black market premium, inflation rate, government size as a share of GDP, indicators of openness of the economy – exports and imports as a share of GDP, indicators of civil liberties, political assassinations, bureaucratic efficiency and corruption.
4 Hypotheses
In the paper I test four hypotheses. First, whether the financial structure, so that distinguishing financial systems into bank and market-based is significant in promoting of economic growth. Second, whether financial development has boosting impact on the level of economic performance, also in case of presence of the variable of financial structure in regression. Third, whether the stage of economic development predicts the structure of the economy (economy being market or bank-based) and whether it has an impact on the level of economic development. Fourth, whether the contractual and legal environment in the economy predicts the specific structure of the economy (economy being market or bank-based) and whether this has an impact on the level of economic performace. Third and fourth hypothesis could be seen as robustness check tests for the relationships obtained for the variables of financial structure.
Four views are taken into consideration when evaluating the results of the analysis. For the bank-based economies, lower values of the variables of financial structure and higher levels of the variables of financial development are typical; therefore should an economy in which banks are prevalent as a source of financing promote economic growth, I would expect the relationship between the variable of financial structure resp. of financial development and economic growth to be negative resp. positive. Empirical evidence is provided by King and Levine (1993) who concludes the positive relationship between banks and long term economic growth in his cross-country analysis for the panel of 80 countries.
14 predominance of the market-based system in the economy and the explanatory power of markets and the market-based view on the level of long-term economic growth.
According to the financial-services view proposed by Levine (1997) is it of significant importance to distinguish between financial systems into the market-based and bank-based ones. From this point of view therefore, I would expect that the variables of overall financial development would be positively related to the level of long-term economic growth.
The law view is supported by La Porta (1997) and La Porta (2000). Whereas La Porta (2000) argues clearly that the differentiation of systems into the ones with the poor and strong enforcement of the legal rights and is crucial. Therefore, from this perspective, I would expect a greater development of systems originating from the English law, common law system than the systems originating from the French law, or civil law-based systems. Furthermore, Rajan and Zingales (1998) argue that in the legal systems where contracts are poorly enforced and laws poorly drafted, bank-based system is more effective and that economies would benefit from the market-based structure only in case of sound legal environment.
Boyd and Smith (1991) state that the growth process is crucial in co-evolution of banks and financial markets. Whereas in the early stage of economic development debt-financed investments can be observable, with an increase of income levels of countries, the position of equity markets becomes competing with the position of banks. In this case I expect that the financial structure variable has a negative effect and level of the legal system development positive effect on the level of long-term economic growth.
5 Methodology
15 It is necessary to mention furthermore, that throught the paper the constructed conglomerate indicators of financial structure and financial development suffer from anomalies, e.g. if a country experiences a high level of financial structure indicators, it does not have to indicate that the economy has a well-developed market system, but it suffers from an underwell-developed bank systems and vice versa. In order to prevent this anomaly, dummy variable which equals 1 is constructed for the countries which have lower than median values of bank credit ratio, market capitalization ratio and total value traded ratio, and higher than median values of overhead costs.
In the second step, in order to estimate the coefficients of financial structure financial development and I insert the constructed conglomerate indicators into the regression for each country for both sample periods and apply OLS method with heteroskedasticity consistent standard errors. In each regression there are 48 observations since the analysis is provided for 48 countries. This small number of observations can possibly lead to unstability of the estimated parameters. Since the aim of this paper is to re-estimate and extend the analysis of Levine (2002), the methodology of this paper is in consistency with Levine (2002), who states in his paper that 48 countries is the maximal number of countries he can obtain data for. Although there are certain rules of thumb concerning the number of observations and number of explanatory variables which enter the regression, Novick et al. (1972), Sawyer and Maxey (1982) and Sawyer (1982) state that eight-variable prediction equations based on representative samples of size 50 would have almost the same accuracy as prediction equations based on larger samples.
In spite of existence of these theories, I provide a stability test on the estimated parameters CUSUM which is derived from the Recursive Least Squares (RLS). The RLS procedure is appropriate only for the time series data and cross-sectional data that have been ordered in the sensible way. CUSUM test is based on normalized version of the cumulative sum of the recursive residuals. Under the null hypothesis of perfect parameter stability, the CUSUM statistic is equal to zero however many residuals are included in the sum (because the expected value of a disturbance is always zero). For CUSUM test it is valid that a set of ±2 standard error bands is usually plotted around zero and any statistic lying outside the bands is taken as evidence of parameter instability. Hence, if the CUSUM line lies within the borders of 5% significance level, the hypothesis of stability cannot be rejected.
In the paper I estimate following estimation formulas:
16 where Gi is the real per capital GDP growth in the form of n×1 vector, X is a set of conditioning information (the simple and the full conditioning information set), S measures the financial structure,
)
1
(
i
is an error term in the form of the n×1 vector;)
2
(
'
i ic
X
dF
G
where Gi is the real per capital GDP growth in the form of n×1 vector, X is a set of conditioning information (the simple and the full conditioning information set), F is an indicator of overall financial development,
i(
2
)
is an error term in the form of n×1 vector;)
3
(
'
i if
X
hS
jF
G
where Gi is the real per capital GDP growth in the form of n×1 vector, X is a set of conditioning information (the simple and the full conditioning information set), F is an indicator of overall financial development and S is a measure of financial structure,
i(
3
)
is an error term in the form of n×1 vector;As stated above, the method applied for the cross-sectional data is OLS (Ordinary least squares method).
If the data consists of n observations {y, i, xi}ni=1.
i i
x
y
where
is ap
×1 vector of unknown parameters;
iis unobserved scalar random variable (error) which accounts for the discrepancy between the actually observedy
i and the fitted outcomesx
i
; and ′ denotes matrix transpose, so that x
is the dot product between the vectors x and
. This model can also be written in matrix notation as
X y
where y and ε are n×1 vectors, and X is an n×p matrix of regressors, which is also sometimes called the design matrix.
17 necessity of data being transformed OLS can be applied directly. In this case, however, the hypothesis testing is more demanding, i.e. there is more evidence necessary to reject the null hypothesis.
In the third step, the alternative approach of evaluating the importance of banks vs. markets in promoting economic growth is tested. This test is provided in line with Boyd and Smith (1998), who state that in the early stage of economic development debt is the major way of financing in the countries and with the increase of the income levels of countries the position of equity markets becomes equal with the position of banks. In other to verify this relationship the interaction term of financial structure and real per capita GDP is included in the following regression:
) 4 ( * 'X bS kS Y U a G
where Y is real per capita GDP, G is the real per capital GDP growth, X is the set of conditioning information, S measures the financial structure, U(4) is the error term.
Another robustness check is provided in line with Rajan and Zingales (1998) who argue that in the legal systems where contracts are poorly enforced and laws poorly drafted, a bank-based system is more effective and economies would benefit from the market-based structure only in case of sound legal environment. In other to control for the fact, that variable of product of financial structure and index of legal system development is included in the regression equation, which has therefore a following form: ) 5 ( * 'X bS kS L U a G
where L is an index of legal system development, G is the real per capital GDP growth, X is the set of conditioning information, S measures the financial structure, U(5) is the error term.
6 Data Sources
In order to re-estimate and extend the analysis of Levine (2002), I firstly run the regression for the initially applied period 1980-1995 and then for the extended period of years 1996-2009. The data for the first sample period is extracted from Levine (2002). In order to keep consistency in the analysis and to be able to construct the comparable results, for the second period I follow the data sources as indicated in Levine (2002). The analysis is provided for 48 countries and appear to be 48 observations in each regression.
19 7 Results
While interpreting the results of the variables of financial structure and development, it has to be born in mind that even though larger values mean the prevalence of the market-based system and lower values indicate that the economy is rather a bank-based one, there is no uniform or strict border given to make a distinction of these systems. Moreover, several anomalies are present when generating the rankings of the variables so the results have to be interpreted carefully. The ranking of the STRUCTURE-ACTIVITY variable indicates that the economies of Taiwan, Malaysia, Switzerland, US and UK are “strongly” market-based. The results of US and UK correspond with the general prediction about these countries being example representatives of market-based economies. The differences between the “degree of being market-based” is relatively high. The value traded ratio of Taiwan is almost four times higher compared to that of US. In case of Switzerland a paradox situation appears on the first sight – the economy is indicated as strongly market-based even though with its bank credit ratio it tops the ranking list. On the other hand, the variable for Turkey, Mexico and Brazil does not reflect the state of the economy correctly. The position as very market-based economy is caused by weakly developed bank system in the economies, not the well developed financial markets. The results correspond with those of Levine (2002).
20 Tobago are classified as strong market-based economies despite having, especially in case of Zimbabwe, extremely underdeveloped bank systems.
STRUCTURE-EFFICIENCY variable indicates that the economies of Switzerland, Taiwan, US, UK and Malaysia have efficient financial markets, therefore are rather market-based of nature. Anomalies arise around Brazil and Turkey which are indicated as the countries with efficient financial systems. It is important to mention however that these countries enter the rankings as market-based since having strongly inefficient banks, represented by high bank overhead ratio, instead of having efficient financial markets. The results for the second sample period classify Switzerland, US, UK, the Netherlands and Taiwan as economies with strongly efficient markets. South Africa enters with high level of the market efficiency in the ranking. The reason lies again in high bank overhead costs not the real high efficiency of the markets.
APPENDIX 2 and 3 present the ranking of conglomerate measures of financial development. Compared to the ranking of the countries according to the financial structure, financial development indicators provide clear and true picture of the economy, not producing anomalies in a big scope. The countries classified as developed ones are US, UK, Switzerland, Taiwan and the Netherlands.
South Africa is indicated as a market with high financial development. However, this may not be true, since this result reflects the presence of large market capitalization in the ratio of the FINANCE-SIZE indicator. However, as already mentioned above, the traded volume in its market is comparatively lower than in the markets with developed financial markets, the variables of financial structure and financial development are most of the time highly correlated with each other. The correlation coefficients are provided in the Appendix.
21 dummies are negative at 1% confidence level, the results remain unchanged.2 The results tend to correspond with the law and finance view, which does not predict financial structure to be a good way to distinguish financial systems in the testing of its impact on economic growth. All variables of interest in the regressions controlling for the simple conditioning information for both periods 1980-1995 and 1996-2009 are proved to be stabile.3
When entering the regression together with overall financial development, the results remain unchanged. Even though, as it will be discussed later, financial development strongly promotes economic development, financial structure does not enter significantly in any case for both sample periods.
TABLE 1
Financial Structure and Economic Growth, 1980-1995, simple and full conditioning information set
The financial structure variables are included one-by-one in the regression (only the variables of interest are shown in the table) for 48 countries. Simple conditioning information set contains the logarithm of initial income (real per capita GDP) and logarithm of schooling. Full conditioning set contains the variables of simple conditioning information set, inflation, black market premium, government size, trade openness, indicators of civil liberties, revolutions and coups, political assassinations, bureaucratic efficiency and corruption. Structure-Activity = ln (total value traded ratio/bank credit ratio), Structure-Size = ln (market capitalization ratio/bank credit ratio), Structure-Efficiency = ln (total value traded ratio*bank overhead ratio), Structure-Aggregate = principal component of the financial structure variables Structure-Activity, Structure-Size, Structure-Efficiency. Dummy variables used in the regressions: OECD dummy, dummy for an african country, dummy for a south asian country, dummy for an arabic country, dummy for a latin american country and structure-dummy – if a country is a bank-based economy it takes 0, 1 is applied for the market-based economy. Similarly as in Levine (2002) a dummy underdeveloped is included in the regression, which equals 1 if the country has below median values of all these financial development indicators. Thus, rather than classifying countries as either bank-based or market-based, I first identify the countries with highly underdeveloped financial systems. In the first panel coefficients of variables of interest controlling for simple conditioning information set are presented, in the below panel, coefficients of variables of interest controlling for simple conditioning information set are presented.
2
In the Table 5 I show the result with inclusion of country dummies. In the other regressions however, I exclude the country dummies completely.
3
The Appendix 7 provides details on the parameter stability analysis. Independent variable
STRUCTURE Coefficient Standard error t-statistic P-value R-squared Activity 0.2311 0.2977 0.7763 0.4436 0.7274 Size 0.0947 0.4374 0.2165 0.8300 0.7223 Efficiency -0.0515 0.2534 -0.2032 0.8403 0.7223 Aggregate 0.1292 0.3942 0.3278 0.7453 0.7229 *, **,*** correspond to significance at 10%, 5%, 1% level
Independent variable
22 This conclusion is also further supported by the results obtained when financial structure enters in the regression together with overall financial development. In approximately half of the regressions, in both controlling for simple and full conditioning set, enters financial structure negatively at 10% confidence level each time. Therefore the results are robustly linked with the bank-based view.
TABLE 2
Financial Structure and Economic Growth, 1996-2009, simple and full conditioning information set
The financial structure variables are included one-by-one in the regression (only the variables of interest are shown in the table) for 48 countries. Simple conditioning information set contains the logarithm of initial income (real per capita GDP) and logarithm of schooling. Full conditioning set contains the variables of simple conditioning information set, inflation, black market premium, government size, trade openness, indicators of civil liberties, revolutions and coups, political assassinations, bureaucratic efficiency and corruption.Structure-Activity = ln (total value traded ratio/bank credit ratio), Structure-Size = ln (market capitalization ratio/bank credit ratio), Structure-Efficiency = ln (total value traded ratio*bank overhead ratio), Structure-Aggregate = principal component of the financial structure variables Structure-Activity, Structure-Size and Structure-Efficiency. Dummy variables used in the regressions: OECD dummy, dummy for an african country, dummy for a south asian country, dummy for an arabic country, dummy for a latin american country and structure-dummy – if a country is a bank-based economy it takes 0, 1 is applied for the market-based economy. Similarly as in Levine (2002) a dummy underdeveloped is included in the regression, which equals 1 if the country has below median values of all these financial development indicators. Thus, rather than classifying countries as either bank-based or market-based, I first identify the countries with highly underdeveloped financial systems. In the first panel coefficients of variables of interest controlling for simple conditioning information set are presented, in the below panel, coefficients of variables of interest controlling for simple conditioning information set are presented.
Table 3 presents the results of the analysis when besides the variable of financial structure, the interaction term consisting of the multiplication of the variable of financial structure and the variables of Income per capita, Shareholder rights vs. Rule of Law enters the first, the second vs. the third regression equations. These tests should be seen as further robustness checks for the conlusions about financial structure.
In the first set of regressions none of the variables enter significantly. The essence of the theory of Boyd and Smith (1996, 1999) lies in the favouring the bank-based system in the early stages of economic development and stressing the supportive role of the market-based system in the further stages. The results of the analysis does not enable to reject the theory in any of the sample periods. The bank-based view predicts negative
Independent variable
STRUCTURE Coefficient Standard error t-statistic P-value R-squared Activity -0.1290 0.2031 -0.6353 0.5308 0.7091 Size -0.5005 0.2798 -1.7882 0.1346 0.6995 Efficiency 0.0396 0.2142 0.1852 0.8544 0.7162 Aggregate 0.0086 0.1662 0.0522 0.9586 0.5248
Independent variable
23 relationship between financial structure and economic performance and the market-based positive one.
TABLE 3
Financial Structure, Interactions with Income, Shareholder Rights and Rule of Law, 1980-1995
The second interaction term and the second regression is based on the theory of Rajan and Zingales (1999) who state that in the weak contractual environment economic growth is promoted more efficiently by the bank-based system taking into account the stronger power of banks to protect their interest compared to financial markets whose efficiency is boosted by the strong contractual and legal environment. Results of the second set of regressions do not prove to be significant in any case. In the the last regression set the variable of law and order tradition of the country is included in the interaction term. This variable can be seen as an alternative to the index of shareholder rights in terms of expressing the contractual environment of a country and lack of respect to the law. However, none of the variables of the financial structure is significantly positively or negatively related to the level of economic growth. The theory of Rajan and Zingales (1999) can not be therefore rejected. Whereas Levine (2002) finds no evidence of significant relationship between any of the financial structure variables for the interaction term including the variables of
The variables of the financial structure are included one-by-one in the regression (only the variables of interest are shown in the table) for 48 countries. In the regression it is controlled for simple conditioning information set, which contains a logarithm of initial income (Real per capita GDP) and logarithm of schooling. The financial structure variables are Structure-Activity = ln (total value traded ratio/bank credit ratio), Size = ln (market capitalization ratio/bank credit ratio), Structure-Efficiency = ln (total value traded ratio*bank overhead ratio), Structure-Aggregate = principal component of the financial structure variables - Structure-Activity, Structure-Size, Structure-Efficiency. Interaction terms consist of S (Variable of financial structure) * Y (Real per capita GDP) in the first regression, Rights – index of equity shareholder rights in the second regression, and LAW – Index of the degree to which the rule of law holds in a country in the last regression). In case of shareholder rights, 1 is added if: (1) the country allows the shareholders to mail their proxy to the firm; (2) shareholders are not required to deposit their shares prior to the General Shareholders’ Meeting; (3) cumulative voting or proportional representation of minorities in the board of directors is allowed; (4) an oppressed minorities mechanism is in place; (5) the minimum percentage of share capital that entitles a shareholder to call for an Extraordinary Shareholders’ meeting is less than or equal to 10 percent (the sample median); or (6) shareholders have preemptive rights that can only be waived by a shareholders’ vote. LAW ranges from 10 (strong law and order tradition) to 1 (weak law and order tradition). The data are averaged over the period 1982-1995. Dummy for law of Scandinavian country, English law, Civil law and French law.
Structure and
Income per Capita Structure and Shareholder Rights
Structure and Rule of Law Independent variables STRUCTURE Coeffic ient P-value Independent variables STRUCTURE Coefficie nt P-value Independent variables Coefficie nt P-value
24 the level of economic growth, the results for he remaining regressions are corresponding with the those ones of Levine (2002).
TABLE 4
Financial Structure, Interactions with Income, Shareholder Rights and Rule of Law, 1996-2009
The variables of the financial structure are included one-by-one in the regression (only the variables of interest are shown in the table) for 48 countries. In the regression it is controlled for simple conditioning information set, which contains a logarithm of initial income (Real per capita GDP) and logarithm of schooling. The financial structure variables are Structure-Activity = ln (total value traded ratio/bank credit ratio), Structure-Size = ln (market capitalization ratio/bank credit ratio), Structure-Efficiency = ln (total value traded ratio*bank overhead ratio), Structure-Aggregate = principal component of the financial structure variables - Structure-Activity, Structure-Size, Structure-Efficiency. Interaction terms consist of S (Variable of financial structure) * Y (Real per capita GDP) in the first regression, Rights – index of equity shareholder rights in the second regression, and LAW – Index of the degree to which the rule of law holds in a country in the last regression). In case of shareholder rights, 1 is added if: (1) the country allows the shareholders to mail their proxy to the firm; (2) shareholders are not required to deposit their shares prior to the General Shareholders’ Meeting; (3) cumulative voting or proportional representation of minorities in the board of directors is allowed; (4) an oppressed minorities mechanism is in place; (5) the minimum percentage of share capital that entitles a shareholder to call for an Extraordinary Shareholders’ meeting is less than or equal to 10 percent (the sample median); or (6) shareholders have preemptive rights that can only be waived by a shareholders’ vote. LAW ranges from 10 (strong law and order tradition) to 1 (weak law and order tradition). The data are averaged over the period 1982-1995. Dummy for law of Scandinavian country, English law, Civil law and French law.
Structure and
Income per Capita Structure and Shareholder Rights
Structure and Rule of Law Independent variables STRUCTURE Coefficie nt P-value Independent variables STRUCTUR E Coeffic ient P-value Independe nt variables STRUCT URE Coefficie nt P-value
Activity -0.1372 0.6857 Activity 0.2936 0.5670 Activity -0.1320 0.7053
Activity*Income 0.0120 0.9611 Activity*Rig
hts -0.1486 0.3580
Activity*
25 TABLE 5
Financial Development and Economic Growth, 1980-1995, simple and full conditioning information set
The financial development variables are included one-by-one in the regression (only the variables of interest are shown in the table) for 48 countries. Simple conditioning informaton set contains the the logarithm of initial income (real per capita GDP) and logarithm of schooling. Full conditioning set contains the variables of simple conditioning information set, inflation, black market premium, government size, trade openness, indicators of civil liberties, revolutions and coups, political assassinations, bureaucratic efficiency and corruption. Finance-Activity = ln (total value traded ratio/private credit ratio), Finance-Size = ln (market capitalization ratio/private credit ratio), Finance-Efficiency = ln (total value traded ratio/bank overhead ratio), Finance-Aggregate = principal componentof the financial development variables Finance-Activity, Finance-Size, Finance-Efficiency . Dummy used in the regression: Finance-dummy – if a country is a bank-based-economy it takes 0, 1 is applied for the market-based economy. Similarly as in Levine (2002) a dummy underdeveloped is included in the regression, which equals 1 if the country has below median values of all of these financial development indicators. Thus, rather than classifying countries as either bank-based or market-based, I first identify those countries with highly underdeveloped financial systems. In the first panel the coefficients of interest using the simple conditioning information set are presented, in the below panel the coefficients of interest using the full conditioning information set are presented.
The second interaction term and the second regression is based on the theory of Rajan and Zingales (1999) who state that in the weak contractual environment economic growth is promoted more efficiently by the bank-based system taking into account the stronger power of banks to protect their interest compared to financial markets whose efficiency is boosted by the strong contractual and legal environment. Results of the second set of regressions do not enter significantly in any case. The incentive to provide this test is provided by the research paper of La Porta (1997, 2000). In the last regression set the variable of law and order tradition of the country is included as the interaction term. This variable can be seen as an alternative one to the index of shareholder rights in terms of expressing the contractual environment of a country and lack of respect to the law. None of the country dummies enters significantly However, similarly as in case of shareholder rights, none of the variables of financial structure is significantly positively or negatively related to the level of economic growth. Therefore, the theory of Rajan and Zingales (1999) cannot be rejected. Apart from the results for interaction term Income per Capita and variable of STRUCTURE-ACTIVITY for the sample period 1980-1995, where Levine (2002) finds no evidence of any significant impact of these on the level of economic growth, results for the sample period 1995-2009 confirm his conclusions about the unimportance of
Independent variable
FINANCE Coefficient
Standard
error t-statistic P-value R-squared Activity 0.5664*** 0.1498 3.7812 0.0005 0.3540 Size 1.0233* 0.5613 1.8229 0.0753 0.2010 Efficiency 0.7215*** 0.1437 5.0184 0.0000 0.3661 Aggregate 1.1944*** 0.3289 3.6317 0.0007 0.3413 Independent variable FINANCE Coefficient Standard
26 division of systems into bank and market-based in any stage of economic development and any contractual or legal environment.
TABLE 6
Financial Development and Economic Growth, 1996-2009, simple and full conditioning information set
The financial development variables are included one-by-one in the regression (only the variables of interest are shown in the table) for 48 countries. Full conditioning set contains the variables of simple conditioning information set, inflation, black market premium, government size, trade openness, indicators of civil liberties, revolutions and coups, political assassinations, bureaucratic efficiency and corruption. Simple conditioning informatoon set contains the the logarithm of initial income (real per capita GDP) and logarithm of schooling. Finance-Activity = ln (total value traded ratio/private credit ratio), Finance-Size = ln (market capitalization ratio/private credit ratio), Finance-Efficiency = ln (total value traded ratio/bank overhead ratio), Finance-Aggregate = principal componentof the financial development variables Activity, Size and Efficiency. Dummy used in the regression: Finance-dummy – if a country is a bank-based-economy it takes 0, 1 is applied for the market-based economy. Similarly as in Levine (2002) a dummy underdeveloped is included in the regression, which equals 1 if the country has below median values of all of these financial development indicators. Thus, rather than classifying countries as either bank-based or market-based, I first identify those countries with highly underdeveloped financial systems. In the first panel the coefficients of interest using the simple conditioning information set are presented, in the below panel the coefficients of interest using the full conditioning information set are presented.
When testing the impact of financial development on the level of economic growth I include, similarly as in case of financial structure, conglomerate variables of financial development by one in the regressions. Table 10 presents the variables of interest which are strongly robustly linked with economic growth. The results for the period 1996-2009 indicate boosting power o the financial development. The results confirm the robustness of the conclusions of Levine (2002) both for the initial period 1980-1995 as well as for the period 1996-2009. All variables of interest in the regressions controlling for the simple conditioning information for both periods 1980-1995 and 1996-2009 are proved to be stabile.4
And finally, Table 7 presents the results of the analysis where both variable of financial structure and variable of financial development enters the regression simultaneously. Each time I include one by one the variable of financial structure and development of the same origin, i.e. financial-activity and structure-activity etc. In the
4
The Appendix 8 provides details on the parameter stability analysis. Independent variable
FINANCE Coefficient Standard error t-statistic P-value R-squared Activity 0.3365*** 0.1021 3.2936 0.0021 0.2554 Size 0.4774** 0.1970 2.4235 0.0200 0.1747 Efficiency 0.4400*** 0.1067 4.1209 0.0002 0.3605 Aggregate 0.5521*** 0.1500 3.6792 0.0006 0.2862
Independent variable
27 analysis for the period 1980-1995, when including the variables of financial structure, financial development continues to have positive impact on the level of economic growth and this relationship is valid at 1% confidence level controlling for the simple conditioning information set. This conclusion is in line with Levine (2002) confirms the consistency with the finance view and does not prove the assumption that financial structure plays an important role in explaining of the level of economic performance. The boosting power of the financial development variables is confirmed also in the sample period 1996-2009. Here it is interesting to mention that again none of the variables of financial structure enters the results positively.
TABLE 7
Overall Financial Development and Economic Growth, 1980-1995, simple and full conditioning information set
The financial development variables are included one-by-one in the regression (only the variables of interest are shown in the table) for 48 countries. Full conditioning set contains the variables of simple conditioning information set, inflation, black market premium, government size, trade openness, indicators of civil liberties, revolutions and coups, political assassinations, bureaucratic efficiency and corruption. Simple conditioning information set contains the the logarithm of initial income (real per capita GDP) and logarithm of schooling. Finance-Activity = ln (total value traded ratio/private credit ratio), Finance-Size = ln (market capitalization ratio/private credit ratio), Finance-Efficiency = ln (total value traded ratio/bank overhead ratio), Finance-Aggregate = principal component of the financial development variables Activity, Size and Efficiency. Dummy used in the regression: Finance-dummy – if a country is a bank-based-economy it takes 0, 1 is applied for the market-based economy. Similarly as in Levine (2002) a dummy underdeveloped is included in the regression, which equals 1 if the country has below median values of all of these financial development indicators. Thus, rather than classifying countries as either bank-based or market-based, I first identify those countries with highly underdeveloped financial systems. In the first panel the coefficients of interest using the simple conditioning information set are presented, in the below panel the coefficients of interest using the full conditioning information set are presented.
Independent variable
FINANCE Coefficient Standard error t-statistic P-value R-squared Activity 0.7966*** 0.1927 4.1330 0.0002 0.3372 Size 1.4489*** 0.4436 3.2662 0.0021 0.213 Efficiency 0.7922*** 0.1729 4.5810 0.0000 0.3741 Aggregate 1.5381*** 0.3427 4.4872 0.0001 0.3454
Independent variable
28 TABLE 8
Overall Financial Development and Economic Growth, 1996-2009, simple and full conditioning information set
The financial development variables are included one-by-one in the regression (only the variables of interest are shown in the table) for 48 countries. Simple conditioning information set contains the the logarithm of initial income (real per capita GDP) and logarithm of schooling. Full conditioning set contains the variables of simple conditioning information set, inflation, black market premium, government size, trade openness, indicators of civil liberties, revolutions and coups, political assassinations, bureaucratic efficiency and corruption. Finance-Activity = ln (total value traded ratio/private credit ratio), Finance-Size = ln (market capitalization ratio/private credit ratio), Finance-Efficiency = ln (total value traded ratio/bank overhead ratio), Finance-Aggregate = principal component of the financial development variables Finance-Activity, Finance-Size and Finance-Efficiency. Dummy used in the regression: Finance-dummy – if a country is a bank-based-economy it takes 0, 1 is applied for the market-based economy. Similarly as in Levine (2002) a dummy underdeveloped is included in the regression, which equals 1 if the country has below median values of all of these financial development indicators. Thus, rather than classifying countries as either bank-based or market-based, I first identify those countries with highly underdeveloped financial systems. In the first panel the coefficients of interest using the simple conditioning information set are presented, in the below panel the coefficients of interest using the full conditioning information set are presented.
8 Conclusions
The paper investigates the impact of financial development and financial structure on economic growth. Several conclusions can be driven from this paper. Firstly, the results of this paper confirm the results of Levine (2002) after re-examining and extending his analysis where strong robust relationship between financial development and the level of economic growth is concluded. In line with Levine (2002), the paper fails to confirm the significant impact of financial structure on the economic performance. Therefore it can be concluded that distinguishing between bank-based and market-based financial systems does not play a significant role in promoting economic growth. The results are consistent with the finance view, which states that financial services have a positive impact on economic growth, with paying little attention to the financial structure itself. These results are in consistence with Tadesse (2000), who provides the analysis on the industry-level data and further with Levine and Zervos (1998) and Beck and Levine (2004). Furthermore, the paper brings no significant evidence of neither the theory of La Porta et al. (1998, 2000) supporting the law view, nor the theory of Rajan and Zingales (1999), nor the theory of Boyd and Smith (1999). In consistence with these papers, the paper confirms the unimportance of financial structure in explaining the level of the economic performance.
Independent variable
FINANCE Coefficient Standard error t-statistic P-value R-squared Activity 0.5718*** 0.1759 3.2497 0.0024 0.3027 Size 0.0045 0.2579 0.0178 0.9859 0.2979 Efficiency 0.8748*** 0.2070 4.2257 0.0001 0.4414 Aggregate 0.8640*** 0.2204 3.9202 0.0003 0.3419
Independent variable
29 Secondly, in the thesis I apply a broad cross-sectional analysis, which is run for two sample periods. Standard method of the research papers is to provide a country-specific time-series analysis applying the country specific measures; in this analysis harmonized measures of financial structure are constructed and used. The obtained variables suffer from several anomalies, but the analysis provides a consistent treatment of financial structure across the countries. This analysis enables me to provide comparisons not only across the countries, but also across the sample periods.
30 References
Arestis, P., Demetriades, P.O., Luintel, K.B., 2001, Financial development and economic growth: The role of stock markets, Journal of Money, Credit, and Banking 33, 16–41;
Arestis, P., and Demetriades, P.O., 1996, Finance and Growth: Institutional Considerations and Causality, Paper presented at the Royal Economic Society Annual Conference, University of Wales Swansea, 1-4 April, 1996;
Arestis, P., Demetriades, P.O., 1997, Financial Development and Economic Growth: Assessing the Evidence,
Economic Journal 107, 783-99;
Beck, T., Levine, R., 2004, Stock markets, banks, and growth: Panel Evidence, Journal of Banking and
Finance 28, 423-442;
Beck, T., Levine, R., Loayza, N., 2000, Finance and the sources of growth, Journal of Financial Economics 58, 261–300;
Bencivenga, V., Smith, S., 1991, Financial intermediation and endogenous growth, Review of Economic
Studies 58(2), 195–209;
Boot, A.W.A., Thakor A.V., 1997, Financial system architecture, Review of Financial Studies, 10 (3), 693-733;
Boyd J.,H., Smith, B.,D. 1996, The Coevolution of the Real and Financial Sectors in the Growth ProcessWorld Bank Economic Review, (1996) 10 (2): 371-396;
Boyd J.,H., Smith, B.,D. 1996, The evolution of debt and equity markets in economic development, Economic Theory, 12 (3), 519-560;
Coffee, J., 2000., Privatization and corporate governance: the lessons from securities market failure.
Working paper no. 158, Columbia Law School, New York;
31 Demetriades, P.O., Hussein, K., 1996, Does Financial Development Cause Economic Growth? Time-series Evidence from 16 Countries, Journal of Development Economics 51, 387-411;
Demetriades, P.O., 2008, New Perspectives on Finance and Growth, Discussion paper no 08/14, University of Leicester;
Gurley, J.G., Shaw E., Financial Aspects of Economic Development, The American Economic Review, 45, No. 4 (Sep., 1955), pp. 515-538;
Hellwig, M., 1998, Banks, Markets, and the Allocation of Risks in an Economy, Journal of Institutional and
Theoretical Economics (JITE) / Zeitschrift für die gesamte taatswissenschaft 154 (1), The New Institutional
Economics Financial Institutions in Transition: Banks and Financial Markets, 328-345;
King, R.G., Levine, R., 1993a, Finance and growth: Schumpeter might be right, Quarterly Journal of
Economics 108, 717–738;
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R. W., 1997, Legal Determinants of External Finance, Journal of Finance 52, 1131 – 1150;
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R. W., 1999, Corporate Ownership Around the World, Journal of Finance 54, 471-517;
Levine, R., 1991, Stock markets, growth and tax policy, Journal of Finance 46, 1445–1465;
Levine, R., 1997, Financial development and economic growth: Views and agenda, Journal of Economic
Literature 35, 688–726;
Levine, R., Zervos, S., 1998, Stock Markets, Banks, and Economic Growth, American Economic Review 88, 537-58;
Levine, R., 2000, Bank-Based or Market-Based Financial Systems: Which Is Better?, Journal of
Financial Intermediation 11, 398–428;
L
evine, R., Beck T., Demirguc-Kunt, A.Financial Institutions and Markets: Across Countries and Over Time32 Novick, M.R., Jacksom, P. H., Thayer, D. T., and Cole, N. S. (1972). Estimating multiple regression in m groups: A cross-validation study. British Journal of Mathematical and Statistical Psychology, 25, 33-50;
Pagano, M., 1993, Financial markets and growth: An overview,
European Economic Review
37, 613-622; Patrick, H. T., 1966, Financial development and economic growth in underdeveloped countries, Economic Development and Cultural Change 14, 174–189;Rajan, R. G., Zingales, L., 1998, The Great Reversals: The Politics of Financial Development in the Twentieth Century,
Journal of Financial Economics
69, 5 – 50;Rajan, R. G., Zingales, L., 1998, Which capitalism? Lessons from the East Asian crisis, Journal of Applied
Corporate Finance, 11 (3), 40–48;
Rousseau, P.L., Wachtel, P., 2000, Equity markets and growth: Cross-country evidence on timing and outcomes, 1980–1995., Journal of Banking and Finance 24, 1933–1957;
Sawyer, R., 1982. Sample size and the accuracy of predictions made from multiple regression equations.
Journal of Educational Statistics, 7(2), 91-104.
Sawyer, R., Maxey, E. J. 1982. The relationship between collge freshman class size and other institutional characteristics and the accuracy of freshman grade predictions (ACT Research Report No. 82). Iowa city, Iowa: The American College Testing Program.
Schumpeter, J. A., 1912, Theorie der Wirtschaftlichen Entwicklung [The theory of economic development] (Dunker & Humblot, Leipzig), translated by Redvers Opie, Cambridge, MA: Harvard University Press, 1934;
Tadesse, S., 2002, Financial Architecture and Economic Performance: International Evidence, Journal of
Financial Intermediation 11, Issue 4, 429-454;
33 APPENDIX
APPENDIX 1
Ranking of the financial structure variables, 1980-1995
Ranking of the financial structure variables throught the period 1980-1995. The analysis is provided for 48 countries. The financial structure variables are Structure-Activity = ln (total value traded ratio/bank credit ratio), Structure-Size = ln (market capitalization ratio/bank credit ratio), Structure-Efficiency = ln (total value traded ratio*bank overhead ratio), Structure-Aggregate = principal component of the financial structure variables - Structure-Activity, Structure-Size, Structure-Efficiency. The bold line indicates the median values.
ACTIVITY SIZE EFFICIENCY AGGREGATE
34
APPENDIX 2
Ranking of the financial structure variables, 1996-2009
Ranking of the financial structure variables throught the period 1996-2009. The analysis is provided for 48 countries. The financial structure variables are Structure-Activity = ln (total value traded ratio/bank credit ratio), Structure-Size = ln (market capitalization ratio/bank credit ratio), Structure-Efficiency = ln (total value traded ratio*bank overhead ratio), Structure-Aggregate = principal component of the financial structure variables - Structure-Activity, Structure-Size, Structure-Efficiency. The bold line indicates the median values.
ACTIVITY SIZE EFFICIENCY AGGREGATE
35
APPENDIX 3
Ranking of the financial development variables, 1980-1995
Ranking of the financial development variables throught the period 1980-1995. The analysis is provided for 48 countries. The financial structure variables are Finance-Activity = ln (total value traded ratio/private credit ratio), Finance-Size = ln (market capitalization ratio/private credit ratio), Finance-Efficiency = ln (total value traded ratio/bank overhead ratio), Finance-Aggregate = principal componentof the financial development variables Finance-Activity, Finance-Size and Finance-Efficiency.
ACTIVITY SIZE EFFICIENCY AGGREGATE
36
APPENDIX 4
Rankings of the financial development variables, 1996-2009
Ranking of the financial development variables throught the period 1996-2009. The analysis is provided for 48 countries. The financial structure variables are Finance-Activity = ln (total value traded ratio/private credit ratio), Finance-Size = ln (market capitalization ratio/private credit ratio), Finance-Efficiency = ln (total value traded ratio/bank overhead ratio), Finance-Aggregate = principal componentof the financial development variables Finance-Activity, Finance-Size and Finance-Efficiency.
ACTIVITY SIZE EFFICIENCY AGGREGATE
