The impact of financial development on economic growth: Evidence from Vietnam

Nội dung text: The impact of financial development on economic growth: Evidence from Vietnam

1. THE IMPACT OF FINANCIAL DEVELOPMENT ON ECONOMIC GROWTH: EVIDENCE FROM VIETNAM * Bui Thi Le Thuy - Nguyen Thi Ngoc Mien 1 ABSTRACT: This paper investigates the effect of financial development to economic growth in Vietnam for the period 1990 – 2017. Using the Autoregressive Distributed Lag (ARDL) Bounds test approach, we find that financial development has a long-run positive impact on the growth of economic. Besides, the error correction model indicates that although there exists the disequilibrium in the relationship between financial development and economic growth in the short-run, the economy will converge back to the long- run equilibrium in the current year. Keywords: financial development; economic growth; ARDL bound test; error correction model. 1. INTRODUCTION In recent year, there has been a vast empirical work on the role of the development of a country’s financial sector to economic growth. However, there is no consensus on the relationship between financial development and economic growth. The first school of thought argues that financial development is the important factors for the growth of the economic by influencing the investment, saving and technological innovations (Beck & Demirguc-Kunt, 2006; Levine & Zervos, 1999). The empirical results of some studies have been advocated this school of thought when they conclude that financial development has the positive effect to economic growth, for instance, Benhabib and Spiegel (2000), Levine et al. (2000), Hermes and Lensink (2013). Another school of thought argues that financial is not the primary source of economic growth and the relationship between financial development and economic growth has been overstressed in literature (Jayaratne & Strahan, 1996; Lucas Jr, 1988). The third view is one of the negative relationship between finance and growth. The high level of liberalization of the financial sector results in decrease the total real credit to the domestic firms, and thereby lowers investments and slows economic growth (Al-Malkawi & Abdullah, 2011; Boyreau-Debray, 2003; De Gregorio & Guidotti, 1995). The finance-growth nexus continues becoming the highly debated issues in financial economics nowadays. According to the Asia Development Outlook reported by Asian Development Bank (2018), Vietnam is expected to continue strong growth with a forecasted GDP growth of 7.1% in 2018 and become one of the three fastest growing economies in the region. As the frontier market and with spurred rapid economic growth, Vietnam presents an interesting case to study. Understanding which factors affecting Vietnamese economic growth is the imperative task. Besides, twenty years after “Doi Moi” policy, the Vietnamese financial sector has developed strongly both in breadth and in depth. Starting from only four State Owned Commercial Banks and a small number of credit cooperatives in 1990, in 2017, the number has increased to 44, including 4 State Owned Commercial Banks, 31 shareholding and joint venture banks and 9 wholly-owned foreign banks. As the result, banking *, School of Economic Mathematics and Statistics, University of Economics Ho Chi Minh City
2. INTERNATIONAL CONFERENCE STARTUP AND INNOVATION NATION 237 networks and services have been expanding rapidly and created great potential for banks to grow their retail banking business. By the year 2017, the capital supply to the economy from the financial system is estimated at 198% of GDP (National Financial Supervisory Commission, 2017). The Vietnamese stock market develops at the high level, attracting lots of foreign investment. In the Vietnamese financial report, the National Financial Supervisory Commission (2016) concludes that the financial system has performed well its function of providing capital for the economy thanks to stable macroeconomics, liquidity of the banking sector and positive developments of the stock market. It is seemed that the financial development accompanies with economic growth, which leads the question whether the financial development has directly impacted to the economic growth. To the best of our knowledge, there is limited study investigating about the directly impact of financial development to economic growth in the case of Vietnam. This study attempts to fill the literature gap by exploring the Vietnamese finance–growth relationship. The rest of the paper proceeds as follows. Section 2 reviews the empirical literature on the relationship between financial development and economic growth. Section 3 outlines describes the set of explanatory variables measuring the financial sector development; some control variables and the dependence variables. In Section 4, we outline the major methodologies utilized in this study. Section 5 reports the results and discussion of results. Finally, Section 6 concludes the paper. 2. LITERATURE REVIEW There is a substantial amount of literature that tries to explore the relationship between financial development and economic growth. Levine et al. (2000) apply the general moment method using a country’s legal origin as the valid instrumental variable to answer whether financial development impact directly to economic growth. Their result shows that financial development is significantly and positively associated with economic growth. Beck et al. (2000) use the variables of banking sector development as the proxy of the financial development. In their study, they conclude that the difference in financial infrastructure across countries has a strong impact on the growth of the economic. A well-developed banking system will improve the legal and accounting standards of the banking sector, which will facilitate financial development and therefore boost the economic growth. Using stock market liquidity and bank credit as the measures of financial development, Beck and Levine (2004) confirm that these two variables are positively correlated with the current and future rates of economic growth. Also using banking system as the proxy for the financial development of the United State, Dehejia and Lleras-Muney (2003) show that well-function banking systems boost economic growth through improving capital allocation. Hermes and Lensink (2013) confirm the importance of domestic financial market when it contribute to mobilizing saving, screen and monitor investment projects, which lead to higher economic growth. Despite of strong evidence of the effect of finance to growth, recent literature begins to question the magnitude and sign of the effect of financial development to economic growth in previous studies. Graff (2003) points out that when economic growth increases at the same rate as that of financial development, there is no causal relationship between them. Besides, when in some case of destabilizing situation such as financial crises, financial development has a negative impact on economic growth. Christopoulos and Tsionas (2004) criticize some previous studies using dynamic panel model to explore the causality running from financial development to economic growth that they ignore the integration and cointegration properties of the data. Therefore, the estimated model can represent a spurious long-run equilibrium relationship. Naceur and Ghazouani (2007) support the idea of no significant relationship between banking and stock market development, and growth. Law and Singh (2014) apply the innovative dynamic panel threshold technique for the data of 87 developed and developing countries and they reveal that more finance is not necessarily
3. 238 HỘI THẢO KHOA HỌC QUỐC TẾ KHỞI NGHIỆP ĐỔI MỚI SÁNG TẠO QUỐC GIA good for economic growth. Samargandi et al. (2014) collect time series data in 42 years for the Saudi Arabia and apply an ARDL bound test to examine the long and short-run impact of finance to economy. The effect of financial development on oil-sector growth is either statistical insignificant or negative and significant. Ductor and Grechyna (2015) use panel data for 101 developed and developing countries over the period 1970 to 2010 and they point out that the effect of financial development on growth becomes negative, if private credit grows rapidly but it does not accompany by growth in real output. For the case of Vietnam, Anwar and Nguyen (2011) apply the generalized method of moments for the data on 61 provinces of Vietnam in the period 1997 to 2006. They use many alternative measures of financial development and find a strong positive link between financial development and economic growth. In our best knowledge, there is no more study investigating about the relationship of Vietnamese finance – growth. 3. VARIABLES AND DATA 3.1. Data description We use annual data covering the period from 1990 to 2017. The Vietnamese annual growth rate of real GDP per capital plays the role as the dependence variable. Some potentially important determinants of economic growth are also collected. The independent variables included in this study consists of annual observations on various measures of financial development, trade openness and the annual average global oil price. We use the principle component analysis to construct a single measure of financial development. The trade openness is measured by imports plus exports to GDP presents the actual status of economic activities within a country. The Brent crude oil spot price – in USD per barrel (Oprice), which is considered as a world benchmark for oil price, is used. 3.2. The construction of financial development variable We collect data on the following three indicators of financial development: The ratio of broad money (M2) to nominal GDP is the first measure to be considered. However, this measure is more related to the ability of the financial system to provide transaction services than its ability to channel the fund from savers to borrowers (Khan & Semlali, 2000). We also use the credit to private sector to nominal GDP as the proxy for financial development, following Beck et al. (2000). Finally, the domestic credit provided by financial sector divided to GDP represents the importance of banking sector, which is the suitable description for financial development (Saci & Holden, 2008). We apply the principal component analysis to construct a single measure of financial development. The main objective of principal component analysis is to decrease the dimensionality in data. According to Ang and McKibbin (2007), this will help to address the problem of multicollinearity or the high correlation between indicators of finance. Second, it cannot be said that any indicator is the best measure of financial development. Therefore, having the summary measure that includes all the financial indicators will provide better information on financial development. Table 1. Principle component analysis Score coefficient matrix Component Eigenvalue Proportion Cumulative Broad Credit to private Credit provided by money sectore financial sector Comp1 2.986 0.995 0.995 0.576 0.578 0.578 Comp2 0.013 0.004 0.999 0.817 -0.432 -0.383 Comp3 0.002 0.001 1.000 0.029 0.693 -0.721 Number of observation = 28
4. INTERNATIONAL CONFERENCE STARTUP AND INNOVATION NATION 239 Table 1 presents the result of principal component analysis with the three of financial development indicators mentioned above. The eigenvalue of the first component is larger than one and explains 99.52% of the standardized variance. Hence, the first component is the best measure of financial development in this case. 4. METHODOLOGY In applied econometrics, there are three commonly used techniques to determine the long-run relationship between series that are non-stationary: Engle and Granger (1987) cointegration technique and Autoregressive Distributed Lag (ARDL) cointegration technique or bounds test of cointegration (Pesaran et al., 2001). In this study, we use the ARDL bounds test to investigate the effect of financial development on economic growth because this technique has some advantages in comparison with other previous methods. First, it is possible to test the cointegrating association between the variables regardless of different orders of integration (Pesaran et al., 2001) while the validity of Engle – Granger and Johansen techniques requires that all the variables be integrated of order one, I(1). The second improvement to the ARDL technique is that it is appropriate to test long-run associations among the series if the sample period is small and it can also correct for probable endogeneity (Pesaran et al., 2001). Moreover, applying the ARDL technique helps us to obtain unbiased estimates of the long-run model (Harris & Sollis, 2003). The ARDL model used in this study is expressed as follows: ∆=+lGDPtββ01 lGDP tt−− 12 + β FD 13 + β lTRD t − 14 + β lOILP t − 1 p qq12 q3 (1) +∆∑γilGDP ti−− + ∑∑ δϕ jtj ∆ FD + k ∆ lTRD tk − + ∑ ηε m ∆ lOILP tmt − + i=1 jk= 11= m= 1 In this equation, lGDP is the natural logarithm of real gross domestic product per capita, FD stands for financial development, lTRD is the natural logarithm of trade openness, lOILP is the natural logarithm of annual average global oil price, and ε is the error term. ββββ1234,,, measure long term relationships. For the purpose of examining the existence of a long- run relationship, we test the null of no long-run cointegration: H01:0βββ= 2 = 3 = β 4 = against the alternative hypothesis which states the presence of a long-run association. We then compare the F-statistic with the critical values (upper and lower bound) given by Pesaran et al. (2001). The null hypothesis is rejected if the F-statistic is above the upper critical value. In this case, the long-run relationship exists between the variables. After testing the relationship among the variables, we estimate the long-run coefficients of the following ARDL model: p qq12q3 lGDPt=+βγ0 ∑∑∑i lGDP ti−− +∆ δ j FD tj + ϕ k lTRD tk − + ∑ η m lOILP tmt − + ε (2) ijk=100= = m= 0 The most appropriate lag length for all the variables are chosen by using Schwarz Bayesian Information criteria. We keep going to estimate the short-run dynamics by using the following error correction model: p qq''12 q'3 ∆lGDPt =β+γ∆01∑ilGDP ti− +δ∆ ∑∑ j FD t − j +ϕ∆ k lTRD tk − +η∆ ∑ m lOILP tm −− +α ECM t +ε t (3) i jk m Finally, we apply number of diagnostic tests to the model to check for the presence of serial correlation, multicollinearity, error in functional form, heteroskedasticity and the stability of long-run coefficients together with the short-run dynamics.
5. 240 HỘI THẢO KHOA HỌC QUỐC TẾ KHỞI NGHIỆP ĐỔI MỚI SÁNG TẠO QUỐC GIA 5. RESULTS AND DISCUSSION 5.1. Unit-root test As discussed previously, the ARDL bounds testing approach can be applied, regardless of the order of integration. The series can be integrated at I(0) or I(1) or I(0)/I(1). However, it must be ensured that none of them are I(2) as the computation process for F-statistics is invalid if this is the case. The ADF unit-root test is used to indicate whether or not the ARDL model should be used and the results are shown in Table 2. Table 2. Unit-root test Variables ADF test In level I(0) First Difference I(1) Intercept Intercept and trend Intercept Intercept a Trend lGDP - 2 . 2 1 9 -2.544 -5.456 -5.416 FD 0.045 -3.019 -6.314 -6.253 lTRD - 0 . 2 1 9 -3.918 -9.195 -8.967 lOILP -0.995 -1.701 -4.386 -4.320 , , * denotes statistical significance at 1%, 5% and 10% level respectively. As can be seen from this table, only lTRD is stationary at the 5% significant level, whereas the others are stationary after first differencing. This implies that using the ARDL bounds test technique to explore the impact of financial development on the economic growth is more appropriate than the Engle – Granger or Johansen methods as we mentioned before. 5.2. Cointegration test Table 3 displays the results from bounds test including the F-statistic, probability and the corresponding outcomes. Based on the bounds test, there is a cointegration relationship among the variables of the first model because the F – statistic, 15.935, is higher than the upper bound critical value at the 1% level of significance. We also have the same outcomes for the second and the fourth model in which FD and lOILP are dependent variables respectively. However, with the third model, the F-statistic from bounds test is just 2,117, less than the lower bound critical value. This means we cannot reject the null hypothesis of no cointegration among those variables. In summary, the result suggests the presence of cointegrating relationship between lGDP and all independent variables. Table 3. Result from Bounds test Dependent Variable F – statistics Probability Outcome FlGDP(lGDP|FD, lTRD, lOILP) 15.935 0.000 Cointegration FFD(FD| lGDP, lTRD, lOILP) 8.956 0.008 Cointegration FlTRD(lTRD|FD, lGDP, lOILP) 2.117 0.484 No Cointegration FlOILP(lOILP|FD, lTRD, lGDP) 7.239 0.020 Cointegration , , * denotes statistical significance at 1%, 5% and 10% level respectively.
6. INTERNATIONAL CONFERENCE STARTUP AND INNOVATION NATION 241 5.3. Long-run impact. Table 4. Estimated long-run coefficients using the ARDL approach ARDL(2, 3, 2, 3) selected based on Schwarz Bayesian Criterion Dependent variable is lGDP Regressor Coefficient Standard error T-ratio Probability Constant 0.8227 * 0.4221 1.95 0.077 FD 0.0069 0.0016 4.43 0.001 lTRD 0.5742 * 0.2772 2.07 0.063 lOILP 0.2251 0.0959 2.35 0.039 , , * denotes statistical significance at 1%, 5% and 10% level respectively. The long-run impact of FD, lTRD and lOILP on lGDP is reported in the Table 4 after estimating the equation (2). From this table, we can see that financial development, the annual average oil price and the trade openess have positive and significant effects on overall economic growth at the 1%, 5% and 10% significance levels respectively 5.4. Short-run impact and adjustment We present the results of estimated coefficients of the error correction model in the Table 5. In the short-run, only oil price is significant at 5% level and has an important positive impact on GDP. Financial development and trade openess have a negative impact but not significant. Besides, the ECM variable has the negative sign. This implies that short-run adjustment, which occurs at a high speed in the negative direction, is statistically significant. This also confirms the cointegration relationship among lGDP, FD, lTRD and lOILP. Moreover, thanks to the value of ECM coefficient, we can see that the disequilibrium caused by the previous year’s shocks dissipates and the economy converges back to the long-run equilibrium. The short-run deviations for the long-run are corrected about 36%. Table 5. Error correction representation for the selected ARDL model ARDL(2, 3, 2, 3) selected based on Schwarz Bayesian Criterion. Dependent variable is ∆lGDP Regressor Coefficient Standard error T-ratio Probability Constant 0.8227 * 0.4221 1.95 0.077 ∆FD -0.0002 0.0005 -0.43 0.673 ∆lTRD -0.0466 0.1249 -0.37 0.716 ∆lOILP 0.0667 0.0275 2.42 0.034 -0.3596 0.0548 -6.57 0.000 ECM()− 1 R2 = 0.9467 , , * denotes statistical significance at 1%, 5% and 10% level respectively. 5.5. Diagnostic test A number of diagnostic tests are applied to the error correction model above. This model passes all the tests against serial correlation (Breusch-Godfrey serial correlation test), heteroskedasticity (White heteroskedasticity test), and normality of errors (Jarque-Bera test). The Ramsey RESET test also suggests that the model is well specified. All the results are summarized in the Table 6 CUSUM and CUSUMSQ stability test results are shown in the Figure 1. The results indicate
7. 242 HỘI THẢO KHOA HỌC QUỐC TẾ KHỞI NGHIỆP ĐỔI MỚI SÁNG TẠO QUỐC GIA the absence of any instability of the coefficients because the plot of the CUSUM and CUSUMSQ statistic fall inside the critical bands of the 5% confidence interval of parameter stability. CUSUM CUSUM squared 1 0 0 CUSUM CUSUM squaredCUSUM 0 6 28 6 28 n n Figure 1. Plot of CUSUM and CUSUMSQ for coefficient stability for ECM model (1) Table 6. ARDL-ECM model diagnostic tests Test statistic Probability Breusch-Godfrey Serial Correlation test 1.237 0.2661 White Heteroskedasticity test 25.00 0.4058 Jarque-Bera test 5.11 0.0779 Ramsey RESET Test (log likelihood ratio) 1.27 0.3477 6. CONCLUSION The core objective of this study is to scrutinize the impact of financial development on the economic growth of Viet Nam by using the recent time series technique of the ARDL procedure. We found that there is a cointegration relationship between overall economic growth, financial development, trade openness and annual average oil price. In the long-run, economic growth is positively affected by financial development. This result is in line with the results of Levine et al (2000), Beck et al. (2000), Beck Levine (2004), Liesa- Muney (2003), Hermes and Lensink (2003) as we mentioned before. Based on these findings, we realize the important role of financial sector in economic development. We hope that the Vietnamese Government will have good strategies to develop the financial system step by step so that it can help to increase the economic growth in the future. REFERENCES Al-Malkawi, H., & Abdullah, N. (2011). Finance-Growth nexus: Evidence from a panel of MENA countries. International research journal of finance and economics, 63(3), 129-139. Ang, J. B., & McKibbin, W. J. (2007). Financial liberalization, financial sector development and growth: evidence from Malaysia. Journal of development Economics, 84(1), 215-233. Anwar, S., & Nguyen, L. P. (2011). Financial development and economic growth in Vietnam. Journal of Economics and Finance, 35(3), 348-360. Asian Development Bank. (2017). Asian Development Outlook 2018: How technology affect jobs. Lightning Source Incorporated. Beck, T., & Demirguc-Kunt, A. (2006). Small and medium-size enterprises: Access to finance as a growth constraint. Journal of Banking & Finance, 30(11), 2931-2943.
8. INTERNATIONAL CONFERENCE STARTUP AND INNOVATION NATION 243 Beck, T., & Levine, R. (2004). Stock markets, banks, and growth: Panel evidence. Journal of Banking & Finance, 28(3), 423-442. Beck, T., Levine, R., & Loayza, N. (2000). Finance and the Sources of Growth. Journal of Financial economics, 58(1- 2), 261-300. Benhabib, J., & Spiegel, M. M. (2000). The role of financial development in growth and investment. Journal of economic growth, 5(4), 341-360. Boyreau-Debray, G. (2003). Financial intermediation and growth: Chinese style: The World Bank. Christopoulos, D. K., & Tsionas, E. G. (2004). Financial development and economic growth: evidence from panel unit root and cointegration tests. Journal of development Economics, 73(1), 55-74. De Gregorio, J., & Guidotti, P. E. (1995). Financial development and economic growth. World development, 23(3), 433-448. Dehejia, R., & Lleras-Muney, A. (2003). Why does financial development matter? The United States from 1900 to 1940. Retrieved from Ductor, L., & Grechyna, D. (2015). Financial development, real sector, and economic growth. International Review of Economics & Finance, 37, 393-405. Engle, R. F., & Granger, C. W. (1987). Co-integration and error correction: representation, estimation, and testing. Econometrica: Journal of the Econometric Society, 251-276. Graff, M. (2003). Financial development and economic growth in corporatist and liberal market economies. Emerging Markets Finance and Trade, 39(2), 47-69. Harris, R., & Sollis, R. (2003). Applied time series modelling and forecasting: Wiley. Hermes, N., & Lensink, R. (2013). Financial development and economic growth: theory and experiences from developing countries: Routledge. Jayaratne, J., & Strahan, P. E. (1996). The finance-growth nexus: Evidence from bank branch deregulation. The Quarterly Journal of Economics, 111(3), 639-670. Khan, M. M. S., & Semlali, M. A. S. (2000). Financial development and economic growth: an overview: International Monetary Fund. Law, S. H., & Singh, N. (2014). Does too much finance harm economic growth? Journal of Banking & Finance, 41, 36-44. Levine, R., Loayza, N., & Beck, T. (2000). Financial intermediation and growth: Causality and causes. Journal of monetary Economics, 46(1), 31-77. Levine, R., & Zervos, S. (1999). Stock markets, banks, and economic growth: The World Bank. Lucas Jr, R. E. (1988). On the mechanics of economic development. Journal of monetary Economics, 22(1), 3-42. National Financial Supervisory Commission. (2017). Macroeconomic report. Naceur, S. B., & Ghazouani, S. (2007). Stock markets, banks, and economic growth: Empirical evidence from the MENA region. Research in International Business and Finance, 21(2), 297-315. Pesaran, M. H., Shin, Y., & Smith, R. J. (2001). Bounds testing approaches to the analysis of level relationships. Journal of Applied Econometrics, 16(3), 289-326. Saci, K., & Holden, K. (2008). Evidence on growth and financial development using principal components. Applied Financial Economics, 18(19), 1549-1560. Samargandi, N., Fidrmuc, J., & Ghosh, S. (2014). Financial development and economic growth in an oil-rich economy: The case of Saudi Arabia. Economic Modelling, 43, 267-278.