Financial Development and Economic Growth: The Case of Greek Economy

This paper examines the relationship among financial development and economic growth, within a framework which also accounts trade openness, for the case of Greece using data covering the period 2001-2017. We investigate this relationship using the Johansen and Juselius (1990) cointegration approach and the Vector Error Correction Models (VECM), employing Granger causality technique, in order to explore the presence of causality among the variables. The results of cointegration analysis suggested that there is one cointegrated vector among the functions of financial development, economic growth and trade openness. Granger causality tests have shown that there are unidirectional causalities running from economic growth to financial development as well as from financial development to trade openness. The results support that financial development and trade openness do not have causal impact on economic growth in Greece, for the aforementioned period. On the other hand, economic growth has a causal impact on trade both directly and indirectly through financial development.

Published by SCHOLINK INC. 1991). Furthermore, these studies emphasize on the positive effect that financial liberalization may play on economic growth by promoting higher savings and higher returns on investments (Galindo, Schiantarelli, & Weiss, 2007).
On the other hand, there are economists who believe that the finance-growth relationship is not important (Lucas, 1988;Chandavarkar, 1992). However, the known of direction of causality remains vital and has important implication for development policy. The relation between financial and economic growth remains unclear.
In recent literature very few studies examine the causality relationship between financial development and economic growth, either in develop or developing economies. In terms of innovative econometric methods and new data, this study investigates the causality relations among financial development and economic growth within a framework that also accounts trade openness for Greece using data over the period 2001-2007. The structure of the paper is as follows: Section 2 briefly presents the literature review. Section 3 presents the data and the econometric methodology. Finally, concluding remarks and policy implications are given in the final section.

Literature Review
In all countries, either developed or developing, the aim of the policy makers is to attain sustainable growth of the economy. The effect of financial development in any process of economic growth of a country has been the subject of numerous studies in the economics and finance literature. Hondroyiannis, Lolos and Papapetrou (2005) examined the relationship between the development of the banking system and the stock market and economic performance for the case of Greece over the period 1986-1999. Their findings suggest that there is a bidirectional causality relationship between finance and growth in the long run. The causality results, using the error correction model, support that both bank and stock market financing can enhance economic growth, in the long run. In addition the contribution of stock market finance to economic growth appears to be substantially smaller compared to bank finance. Yucel (2009)  Granger causality results suggest that trade openness and financial development can affect economic growth in the country. Furthermore economic growth has causal impacts on trade and finance implying the support for growth-led trade hypothesis but not the trade-led growth model. Rachdi and Mbarek (2011) examined the direction of causality between finance and growth for a sample of 10 countries, 6 from the OECD region and 4 from the MENA region during 1990-2006.
Their empirical analysis confirms a long-term relationship between financial development and economic growth for the OECD and the MENA countries. Findings show that financial development and real GDP per capita are positively and strongly linked. Finally, the causality analysis shows the existence of a bidirectional relationship for the OECD countries and of a unidirectional causality running from economic growth to financial development for the MENA countries. Kaushal and Pathak (2015) investigated the causal relationship among financial development, economic growth and trade openness in India for the post liberalization period ranging from 1991-2013.
Their findings suggest that economic growth and financial development have a positive effect on trade openness. The results recommend that India should consider economic policies which support the philosophy of growth-led trade, where dependence on foreign direct investment might be a feasible option.

Data
The  Table 1.

Methodology
The relationship between financial development, economic growth and trade openness can be expressed as follows (see also the study of Kaushal and Patahk 2015): where t  is the white noise.
After descriptive statistics, this papers uses unit root techniques to examine the stationarity of the three variables and then cointegration approach to investigate the long run relationship among them. Finally, a dynamic panel Vector Error Correction Model (VECM) is used in order to find the short and long run Granger causal relationships between financial development, economic growth and trade openness in Greece.

Unit Root Tests
The literature proposes several methods for unit root tests. Since these methods may give different results, we selected ADF by Dickey-Fuller (1979), PP by Phillips-Perron (1988) and DF-GLS by Elliott, Rothenberg and Stock (1996). In all these tests, the null hypothesis is that the variable contains a unit root (i.e., it is not stationary).

Cointegration Analysis
Since unit root tests have been applied, we continue by testing the long run relationships between financial development, economic growth and trade openness for the Greek economy, using the Johansen and Juselius (1990) cointegration approach. Johansen and Juselius have (1990) developed two tests to detect the number of cointegrating vectors: the maximum-likelihood test and the trace test.

Vector Error Correction Models
Once the variables are proved to be cointegrated, two different kinds of equations arise: i) The long-run equation: where FD, GDP and TO represent financial development, economic growth and trade openness, respectively. In addition t u is the stochastic error term with mean zero and a constant variance.
ii) The short-run model or the vector error-correction representations: where i (i=1,…p) is the optimal lag length determined by the Akaike information criterion (AIC), where ECM t-1 stands for the lagged error correction term from the long-run cointegration equation (Eq. 2), λ 1 , λ 2 , λ 3 are the adjustment coefficients, and ε 1t , ε 2t , ε 3t are the disturbance terms assumed to be uncorrelated with zero means N(0,σ).

Unit Root Results
We begin applying the unit root tests of ADF (of Dickey-Fuller 1979), PP (of Phillips-Perron 1988) and DF-GLS (of Elliott, Rothenberg & Stock, 1996). The results of level and first difference unit root tests for the three variables are provided in Table 2. As can be seen from Table 2, the results showed that all variables (FD, GDP, TO) contain a unit root (non-stationary) in levels. In all cases the tests confirm the stationarity hypothesis, either with intercept or including intercept and trend. Evidently, the results indicated that all variables are stationary in their first differences (i.e., I(1)).

Cointegration Results
After identifying the order of integration, we then use the Johansen and Juselius (1990) Full Information Maximum Likelihood (ML) technique to investigate cointegration for long run relationship between the examined variables. Akaike Information Criterion (AIC) was used to determine the optimum lag length selection, while maximum lag length is set up to level four. The results of the Johansen and Juselius's cointegration test are presented in Table 3.  Table 3 show that both the maximum eigen value and trace tests statistics have their values greater than the critical values at 5 percent level of significance. Therefore, the null hypotheses of no cointegrating vectors (r=0; r≤1) against the specific alternatives are clearly rejected.
Thus it is possible to say that there are long run equilibrium relations between three variables.
The cointegrating vector is shown below: (0.000) (59.55) (standard error in parentheses) The above equation shows that if the GDP increases by 1% then there is a growth in FD of 0.28 % and if TO increases by 1% there is an increase in FD of 1.83%.

Granger Causality Test Based on VECM
In order to investigate the short and long run dynamic relationships among the variables of financial development, economic growth and trade openness we adopt the two steps Engle and Granger (1987) method. The existence of cointegration between the examined variables implies that there is causality relation among them in at least one direction (Engle & Granger, 1987). However, the direction of www.scholink.org/ojs/index.php/ijafs where i (i=1,…p) is the optimal lag length determined by the Schwarz Information Criterion (SIC), ECM t-1 is the lagged residual obtained from the long-run relationship presented in equation, λ 1 , λ 2 , λ 3 are the adjustment coefficients, and u 1t , u 2t , u 3t are the disturbance terms assumed to be uncorrelated with zero means N(0,σ). Notes. Δ denotes first difference operator. ** and * significant at 5% and 10% levels. Short-run causality is determined by the statistical significance of the partial F-statistics associated with the right hand side variables. Long-run causality is revealed by the statistical significance of the respective error correction terms using a t-test.
From the results of Table 4 we can see that: There are two short run unidirectional causalities running from GDP to FD as well as from FD and TO.
We can point out that, according to the result, trade openness is affected both by financial development (directly) and economic growth (indirectly through FD, see Figure 1). In the long run, the estimated coefficients of ECT in equations of financial development and trade openness are negative and statistically significant at 5% level, implying that financial development and trade openness could play an important adjustment role as the system departs from the long-run equilibrium.

Conclusion and Policy Implications
This study investigates the relationship between financial development and economic growth within a framework which also accounts trade openness in Greece using the Jojansen's maximum likelihood procedure in a multivariate model over the period 2001-2017.
Findings suggest that there is a strong evidence of cointegration between the three variables, which indicates that there is a long-run equilibrium relationship. The cointegration relationship indicates that an increase 1% of economic growth has as a result an increase 0.28% of financial development. In addition, increase 1% of trade openness will cause an increase of 1.83 in financial development.
The causality results based on the Vector Error Correction Model (VECM) show both in the short and long run that financial development and trade openness do not have casual impact on economic growth.
On the other hand, economic growth has a casual impact on trade both directly and indirectly through financial development. Findings support the growth-led trade hypothesis instead of trade-led growth.
This follows the studies of Soukhakian (2007) in Japan,  in Nigeria and Kaushal and Pathal (2015) in Indian.
The new model in the theory of growth considers technological progress as an endogenous factor and foreign direct investments to have a permanent effect on the development through technology transfer.
Therefore, Greece should immediately implement policies to attract foreign direct investments and foreign capitals in order to promote economic growth and enhance financial and trade liberalization.