Financial structure and economic growth

Financial Structure And Economic Growth NGUYEN THI UYEN UYEN University of Economics HCMC - uyentcdn@ueh.edu.vn TU THI KIM THOA University of Economics HCMC - tkthoa@ueh.edu.vn PHAM TH

Financial Structure And Economic Growth

NGUYEN THI UYEN UYEN

University of Economics HCMC - uyentcdn@ueh.edu.vn

TU THI KIM THOA

University of Economics HCMC - tkthoa@ueh.edu.vn

PHAM THIEN BACH

University of Economics HCMC - bachphamthien@gmail.com

Abstract

Using the FMOLS method on a data sample of six Asian countries provided by World Bank, this paper aims

at examining the effects of financial structure and development on a country’s economy and also economic growth Our results show that financial structure and development do affect a country’s economy and economic growth There also exists a diversity in countries’ suitability with each type of financial structure South Korea and Malaysia suit a bank-based financial system, whereas Vietnam suits a market-based financial system, and Thailand favors a neutral system, benefiting from the growth of financial services’ growth in quality and quantity Besides the main estimation function, the results of additional estimation functions also support our original conclusion

Keywords: financial structure; economic growth; Asian countries

1 Introduction

Whether a country’s financial system affects its economic growth remains a controversial topic among researchers In general, there exist four schools holding different views on this subject; they are the bank-based school, market-based school, financial services/neutral school, and law and finance school The bank-based school upholds banks’ (both commercial and state-owned) role in stimulating economic growth In contrast, the market-based school argues for a strong and efficient financial market with excellent liquidity This school believes that such a market would stimulate overall growth, increase company efficiency, diversify, and optimize risk management tools, which

in turn makes it easier for companies to maximize their asset value The neutral school, however, disagrees with the other two Those belong to this school dismiss the emphasis on financial structure’s importance, on the contrary, they believe that it is the quality and quantity of financial services provided by the financial market that affects economic growth The last school, the law and finance school, believes in the utmost importance of overall orderliness of a financial market as well

as upholding the underlying financial law Each school has its own advocates and supporting empirical evidence, providing a perfect research environment and opportunity Therefore, this paper seeks to determine the true effects of financial structure and development on economic growth as our primary concern We also set out to examine the possible suitability difference among countries with different economic backgrounds as well as different average income Furthermore, we also test the effects of financial liberation and development on Vietnam’s economy Since the “Doi Moi” plan effectively kicked in 1986, the Vietnam government has imposed many important economic liberalization policies Two of those are the opening of the Vietnam’s stock market in July 28, 2000 and the emergence of commercial banks in the 1990s The Vietnamese financial market has since flourished, unlocking hitherto unknown potentials and helping new businesses fund their operations

In order to achieve our academic goals, we employed the FMOLS method first introduced by Pedroni (2001), which has been proven very effective in examining small size sample Our estimation results showed that both the financial system as a whole and the very nature of a country's financial system

do play an important role in promoting economic growth in most countries in our sample

To achieve our goals, this paper will focus on these primary questions:

First and foremost, is there a direct, one-way relationship between financial structure and

development and economic growth?

Second, does the suitability of countries’ economy to different types of financial system shift when

such countries achieve a higher level of economic development?

2 Literature summary

2.1 Benefits of a sound financial system

One of the most important roles of a financial system is that of a capital channel The amount of available and liquid capital in an economy plays a crucial role in determining said economy’s prosperity Free and private capital in an economy mostly comes from the remainder after spending

An ideal financial system will collect this amount of capital, transform it into working capital by issuing standardized financial instruments to help capital owners invest and earn back interest, as well as diversify their investments A country’s financial system can also transform low liquidity assets into high liquidity ones With these high liquidity assets, investors as well as debtors can hold financial instruments and convert them back to money if they need Financial systems also help investors to diversify their investments and reduce information asymmetries An efficient financial system will also eliminate “friction” or costs that investors burden themselves with in order to obtain

a well-diversified portfolio

2.1.1 Bank-based school

The bank-based school represented by Gerchenskron (1962); Diamond (1984); Stiglitz (1985); Boyd and Prescott (1986); Bencivenga and Smith (1991); Bhide (1993) and Stulz (2002) all uphold the importance of banks in stimulating the economic growth In a bank-based country, the health of said country’s economy depends a lot on the health of its banking system Banks’ profit mostly comes from lending interest which is also the source of their financial power. Banks in this type of financial system scarcely have to deal with competition from the rest of the market Therefore, companies and new businesses do not have many options to fund their operation other than accepting banks’ services Some studies suggest a link between a country’s level of economic development and its suitability to a specific type of financial system Specifically, a country with a lower level of economic development will favor a bank-based financial system and vice versa The reason for this lies in the need for personal investments of the private sector In general, financial markets, especially stock markets which have always been infamously risky does not exactly match the risk profile and the taste of people with no to little knowledge in finance and not a lot of reserve capital

2.1.2 Market-based school

The market-based school represented by the likes of Jansen and Murphy (1990); Holstrom and Tirole (1993); Boot and Thakor (1997); Levine (1997); Boyd and Smith (1998); Wenger and Kaserer (1998) instead favors a large and efficient financial market in the role of economy stimulation As the name would have suggested, in this type of financial system, power lies within the market Strangely, more banks would be involved in the day-to-day activities of the market; however, banks’ profit mostly comes from underwriting activities rather than the interests from loans For obvious reasons,

a market-based financial system will be favored over a bank-based one when a country achieves higher levels of economic development

2.1.3 Neutral school

Unlike the aforementioned two schools, scholars belong to this school simply think that a sound financial system as a whole plays a crucial role in stimulating growth They differ significantly thanks

to their indifference view on the matter of favoring the banking system or the financial market There

is surely a type of pragmatism to this school Scholars belong to this school advocate strongly for expanding the scope and the scale of the financial market, improving the quality, and increasing the quantity of financial services available

2.2 Literature review

One of the first papers seriously debating this subject is Demetriades (1996), which used the VAR method with appropriate cointegration tests on a data set of carefully selected and filtered countries was unable to find a direct relationship between financial and economic development Christopoulos (2004) applied the FMOLS (fully modified OLS) first proposed by Pedroni (2001) on a data set composed of countries with different economic backgrounds and found contrasting results Duisenberg (2001) criticized the need to distinguish between a bank-based and market-based financial system and advocated strongly for a cooperating relationship between these two factors in

a financial system According to Duisenberg (2001), scholars should only care about the nature of financial services provided by these two factors Luintel and Khan (1999) also applied the VAR model

on a data set composed of 10 countries; however, this paper differs greatly from its precedents by considering, not discarding the fundamental political and historical differences between countries Calderon (2002) used the Geweke analysis method on a comprehensive panel data set of 109 countries from 1960 to 1994 and found evidence for a direct causal relationship between financial development and economic growth Levine (2002), nevertheless, failed to find such a relationship in his study of 48 countries from 1980 to 1985 when all the obtained results indicate that financial development could not explain economic progress in these countries This result is mirrored by that

of Beck and Levine (2002) which constructed a coherent data set based on financial and economic data of 36 industries in 42 countries into perhaps the most comprehensive variable set measuring both economic and financial development They found that financial development cannot explain neither overall economic growth nor other measures such as specific industry growth or the number

of new businesses While examining a data set of 41 countries from 1976 to 1993, Levine and Zevros (1996a, 1996b) found supporting evidence for their claim that overall liquidity in financial markets promote higher levels of capital accumulation and productivity which ultimately lead to higher economic prosperity Rajan and Zingales (1998) came to the same conclusion Arestis et al (2001) through their research concluded that the health of a country’s financial market is crucial to said country’s economy Choe (1999) closely examined the effects of both the financial market and the banking system of South Korea on its booming economy through the 1970s period and came to the conclusion that in South Korea, a strong banking system is more favorable but also admitted that such a system may only be suitable to South Korea with its unique economic and historic characteristics Choe (1999) is also one of the few papers which can demonstrate via econometric methods that the supposed superiority of one system over another, in this case, is the banking

system Bencivenga (1991) found evidence supporting the mere existence of banks in an economy in general and in a financial market specifically can be beneficial by easing the flow of capital

3 Model and estimation methods

3.1 Model

Based on the legacy of Luintel et al (2007), we regress countries’ GDP per capita on per capita fixed investment as well as financial structure and financial system development using the following function:

log (Q/L) t = a 0 + a 1 log (K/L) t + a 2 log (F S ) t + a 3 log (F D ) t + e 1 (1)

in which, Q is GDP, L is the total labor force or population, K is the amount of fixed investment,

FS and FD in order are variables representing the financial structure and the degree of financial development, and e1 is the error term

During the regression, we use real GPD per capita (YP) and real fixed investment per capita (KP) Based on the way these two variables were constructed, a higher value of FS will indicate a more market-based financial system; a higher value of FD will indicate a more developed financial system Function (1) therefore will be our main estimation function We care more about the statistically significance of a2 and a3 than their signs since a statistically significant a2 will confirm a country’s financial system’s effect on its economy Theoretically, the market-based school predicts a positive and statistically significant a2; in contrast, the bank-based school predicts a negative and statistically significant a2 The neutral school on the other hand predicts a statistically insignificant a2 All three schools expect a3 to be positive and statistically significant

3.2 Data and variable description

3.2.1 Data description

The data used in this paper is economic and financial data of six Asian countries which are South Korea, Malaysia, Thailand, India, Philippines, and Vietnam with different economic development backgrounds Economic data composed of data on total GDP, total GFI (gross fixed investment), inflation rate, and total population is obtained through the selection of available data provided by World Bank website Specialized financial data composed of data on Stock Market Capitalization, Total Stock Value Traded, Total Stock Market Return and Private Credit Ratio is provided by World Bank’s Financial Development and Structure Dataset Since we do not intend to pool data from different countries into one set of panel data, instead we use FMOLS to run a time series regression, the difference between countries’ data timeline is not a concern Most time series in our sample has

a range from 1989 to 2014, except for Vietnam’s case since its stock market was not officially opened until 2000 and financial data took four years to be truly reliable, Vietnam’s time series start from

2004 and end in 2014

3.2.2 Variable description

3.2.2.1 Dependent variable – GDP per capita (YP)

We deflate nominal GDP per capita to achieve real GDP per capita and then divide it by the total population An increase in YP signals economic growth

Y P = Real GDP

Total Population

3.2.2.2 Independent variables

- Physical capital stock per capita (K/L)P

We deflate nominal GFI data obtained (?) through World Bank and then divide it by the total population For obvious reasons, we expect this variable’s coefficient to be statistically significant and positive

(K/L) P = Real GFI

Total Population

- Financial structure – FS

Two variables used to describe a country’s financial system are: (1) Structure Activity – SA; and (2) Structure Size (SZ) Structure Activity is calculated as the logarithm of the ratio of Total Stock Value Traded to Total Private Credit This variable is a proxy to measure the relative activity of the stock market compared to the banking system The correct measurement of this variable is crucial, which arises from the need to distinguish between two unrelated notions: stock market activity versus stock market size Since a big stock market with many listed companies can still be a ghost market with little to no trading activity Structure Size is calculated as the logarithm of the ratio between Total Stock Market Capitalization and Total Private Credit This variable is a proxy to measure stock market’s size compared to the rest of the financial sector FS representing financial structure as a whole is the weighted average of all the principal components of SA and SZ, therefore being able to capture the variations in financial structure This is an important independent variable since its coefficient’s characteristics will determine the relationship between financial structure and economic prowess Since a higher level of FS signifies a more market oriented financial system, the bank-based school expects its coefficient to be negative and statistically significant, while the market-based school expects its coefficient to be positive and statistically significant On the other hand, the neutral school expects FS’s coefficient to be statistically insignificant

- Financial development - FD

Two variables used to measure the level of financial development are: (1) Financial Size (FZ); and (2) Financial Activity (FA) FZ is calculated as the logarithm of the product between Total Private Credit and Total Stock Market Capitalization This variable measures the total size of the financial market as a whole FA is calculated as the logarithm of the product between Total Private Credit and Total Stock Value Traded This variable measures the level of activity of the financial system FD is

the weighted average of all the principal components of FZ and FA All aforementioned schools predict a positive and statistically significant FD

.

3.2.2.3 Construction variables

- Total Private Credit

This variable is calculated as the ratio between the total amount of credit possessed by the financial system which includes banks as well as private financial entities and GDP

- Total Stock Market Capitalization

This variable is calculated as the ratio between the total stock market capitalization value and GDP

- Total Stock Market Turnover

This variable is calculated as the ratio between the total stock value traded and average stock market capitalization for the period

- Total Stock Value Traded

This variable is calculated as the ratio between the total stock value traded and GDP

Table 1 gives an overall description of variables, and Table 2 presents their calculation formulas

We chose six Asian countries with different economic, political, and historical backgrounds Vietnam has the lowest level of GPD per capita (68.523 USD), while South Korea has the highest one (9959.447 USD) based on the 1989 USD Malaysia has the fastest average GDP per capita growth rate (3.1%), while Vietnam even has negative average GDP per capita growth (-1.6%), we also recorded many instances of negative GDP per capita growth in India and Philippines However, we do notice a sample wide trend of leaning towards a more market oriented financial system By observing an ever more increasing PC or specifically PCg of all countries except for India and Philippines, albeit with different degrees in each country Interestingly, Vietnam is the only country in our sample with double-digit average growth in Total Private Credit (11.3%) This has demonstrated the intensive changes made

to Vietnam’s financial market and their results Vietnam’s average growth rate in Total Private Credit also dwarfs other countries’ However, the average growth rate of Total Stock Market Capitalization and Total Stock Value Traded are far more pronounced This trend is presented in all countries in our sample

Table 1

Variable summary

SZ m SZ g FZ s FZ e FZ m FZ g FA s FA e FA m FA g

Notes: Yg is the average annual growth rate of GDP per capita PC = Private Credit by Deposit Money Banks and Other Financial Institutions to GDP ratio SM = Stock Market Capitalization

to GDP ratio SA = Stock Market Total Value Traded to Private Credit by Deposit Money Banks and Other Financial Institutions SZ = Stock Market Capitalization to Private Credit by Deposit Money Banks and Other Financial Institutions Subscripts S, E, M and G are the average value of the first five years, the average value of the last five years, the mean value of the sample period and the average annual growth rate respectively

Table 2

Variables definition

Total Population

Total Population

GDP

GDP

Average Market Capitalization for the period

GDP

𝑃𝐶 )

𝑃𝐶 )

4 Results

ADF tests with our independent variables show that all of them are all unit root processes Therefore, we regress the logarithm of these variables instead Table 3 presents the regression results: