. International Journal of Economics and Financial Issues ISSN: 2146-4138 available at http: www.econjournals.com International Journal of Economics and Financial Issues, 2017, 7(3), 196-209. International Journal of Economics and Financial Issues | Vol 7 • Issue 3 • 2017196 Financial Development and Economic Growth: The Empirical Evidence of the Southern Mediterranean Countries Mohamed Aydi1, Abdelkader Aguir2* 1Faculty of Economic Sciences and Management of Sousse, Tunisia, 2BETA Lab UMR 7522 University of Lorraine and UR MOFID UR 13-ES60, France. *Email: abdelkader.aguir@univ-lorraine.fr ABSTRACT This paper analyzes empirically the links between financial development and the economic growth of the (squared multiple correlation). The study is based on a vector autoregression approach: The Johansen tests for cointegration and vector error correction model models. The debate on the relation between the financial sphere and the real economic sphere was very ambiguous some studies have shown a positive association between these two spheres while others presented the opposing view perfectly. On the basis of the data relative to the PSM, observed during period 1981-2014, we tried to show there is or not a positive relationship between the financial development and the economic growth. This relationship and increasingly intense for the role of the banking development and and refers to an innovation effort and modernization of the financial system. Keywords: Financial Development, Economic Growth, Southern Mediterranean Countries, Vector Autoregression Approch JEL Classifications: G10, O47 1. INTRODUCTION The theoretical debates on the sense of causality between the financial development and the economic growth, were marked by a significant advance progress. Two currents of the literature come to intervene: One shows the favorable effect of the development of the banking sector and the financial market on the economic growth, while the other supports the opposite view perfectly. The actions of openings and revitalizing the financial system generally and the banking system in particular are causing financial instability and spread of banking crises which were translated by a decline of the economic growth which is due to the importance of the envisaged costs. The positive effect of financial development on economic growth was initially studied by the authors of the school of financial repression namely McKinnon (1973) and Shaw (1973) and the authors of the liberal school namely Keynes and Hicks. These authors showed that an efficient financial system, dynamic and renovated is at the; origin of a capital accumulation, of a stimulation of the investment and then for economic development. The adverse effect of development of the banking system and the financial market on the economic growth was derived from recent banking and financial crises in the context of a financial liberalization policy. On the one hand, the strong information asymmetry characterizing the financial markets, may be the cause of a unsuccessful in coordinating the allocation of savings to investment. This information asymmetry companion can distort investors’ expectations who prefer to invest in less risky than in another universe uncertain and risky. This taking into account the investor’s degree of risk aversion, imperfect financial markets and high transaction costs. This dysfunction of the financial market and the inefficient intermediation can only slow economic growth. Moreover, the recent crises of banking insolvency have plunged economies for the periods of recession. This experience gave us an example of the negative influence of the development of the banking sector onto the macroeconomic performance. These banking dysfunctions can be transformed into banking or financial crises generating huge costs for the whole economy. The absence of consensus regarding the effect of the development of banking and financial market on economic growth brings us Aydi and Aguir: Financial Development and Economic Growth: The Empirical Evidence of the Southern Mediterranean Countries International Journal of Economics and Financial Issues | Vol 7 • Issue 3 • 2017 197 to verify this relation for Tunisia. To address this problem, a theoretical study and an empirical validation appear to be useful. To do this, we propose to use the time series methods based on unit root tests and cointegration Johansen and the Granger causality test. The advantage of cointegration test is the detection of a stable long-term relationship between financial development and economic growth. The rest of the article is organized as follows, section 2 reviews the literature. Then, the section 3 describes the data and defines the variables used. Then, the section 4 exposes the methodology, followed by the presentation of the results in the section 5. Finally, the section 6 is reserved for the conclusion and for the implications of economic policy. 2. LITERATURE REVIEW Several studies (cross-sectional data, panel data and time series) have focused on the nature of the relationship between financial development and economic growth. The results of these studies depend essentially on the nature of the selected sample. Table 1 shows the chronological list of empirical studies on time series, which demonstrated the link between finance-growth. 2.1. Summary of the Main Empirical Research We are interested in the review of the empirical literature to work on time series because this article is based on a technique of time series of PSM countries to study the impact of financial development on the economic growth. Research on the time series are particularly relevant when we want to estimate the sense of the causality between the financial development and the economic growth. The first empirical works of time series reported to Gupta (1984) and Jung (1986) uses the Granger causality tests and the VAR model “vector auto regressive” in level. By estimating the specification of a vector error correction, Demetriades and Hussein (1996) tests Demetriades and Hussein (1996) tests the link of long-term causality between the development of the financial intermediation and the growth of the real gross domestic product (GDP) per capita, respectively 16 and 10 developing countries. These authors support strongly the presence of a bidirectional causality and the existence of a reverse causality, very low, from the growth to the financial development, with results highly varied between economies. Luintel and Khan (1999) detects on the contrary a bidirectional connection between the financial development and the economic growth of all the countries of the sample. They explain the gap with the results of Demetriades and Hussein (1996) used longer time series and use a multivariate approach (rather than two variables). The presence of a bidirectional causality between the financial development and the growth, for developing economies, is questioned by Xu (2000). Xu (2000) on a sample of 41 developing countries between 1960 and 1993 demonstrated the presence of a positive effect of financial development in the long term, but short term is unfavorable, on the economic performance of most developing countries. It uses a multivariate approach VAR that allows the identification of long- term cumulative effects of financial development on GDP growth and the effects of the investment, by taking into account dynamic, short-term interactions, between variables. Focusing on the case study of Malaysia between 1960 and 2001, Ang and McKibbin (2007) indicated that, contrary to the results (profits) obtained by Xu (2000), this growth is at the origin of the development of the long-term banking sector, and not the opposite. Similar conclusions are established by Abu-Bader and Abu-Qarn (2006) for a sample of five countries in the MENA region, between 1960 and 2004. These authors show that the long-term relationship establishes between the financial development and the growth for the economies of these countries, is either bidirectional, or going from growth to the development of the financial system. Ozturk (2008) reviewed the literature on finance-growth nexus and investigate the causality between financial development and economic growth in Turkey for the period 1975-2004. The empirical findings in the paper show a two way causality (bidirectional) between financial development and economic growth. The existence of causality in both sensesd between the finance and the growth is rarely validated for the case of the developed countries. Based on the error correction model Rousseau and Wachtel (1998) show that the dominant direction of the long-term causality in 5 industrialized countries studied, is the one part of the financial development to economic growth, not the reverse. Through a vector error correction model (VECM) analysis Arestis et al. (2001) obtain the same conclusion for the same group of countries (5 industrialized countries), after the integration of stock market development indicators. They show, besides, that it is the banks who contribute most (the most significant and the most important) to the process of growth in these countries compared to the stock market (for two of five studied countries, the effect of the development of the stock market on the growth is negative). Neusser and Kugler (1998) confirm these results by the application of Granger and Lin causality test (1995) for a sample of 13 countries of the OECD between 1970 and 1991. These authors estimate the relation finance - growth by penetrating of two variables via, respectively, into the GDP of the financial system and into the made GDP. Having analyzed the empirical work to defend the presence of a linear relationship between financial development and growth, we try to apply the techniques of time series on our samples. Acaravci et al. (2009) review the literature on the finance-growth nexus and investigate the causality between financial development and economic growth in sub-Saharan Africa for the period 1975-2005. Using panel co- integration and panel GMM estimation for causality, the results of the panel co-integration analysis provide evidence of no long-run relationship between financial development and economic growth. The empirical findings in the paper show a bi-directional causal relationship between the growth of real GDP per capita and the domestic credit provided by the banking sector for the panels of 24 sub-Saharan African countries. The findings imply that African countries can accelerate their economic growth by improving their financial systems and vice versa. Aydi and Aguir: Financial Development and Economic Growth: The Empirical Evidence of the Southern Mediterranean Countries International Journal of Economics and Financial Issues | Vol 7 • Issue 3 • 2017198 Authors Number of the country(s) Period Methodology Variables The empirical results Studies on cross-sectional data Goldsmith (1969) 35 countries Annual data between 1949 and 1963 OLS and graphical analysis The variables of financial development and economic growth The presence of a positive relationship - albeit statistically weak - between financial development and growth Atje and Jovanovic (1993) 94 countries Annual data between 1960 and 1985 MCO The variables of development of stock markets and economic activity Significant positive effect of the development of stock markets on the level and the growth of the economic activity Harris (1997) 39 countries Annual data between 1980 and 1988 Double least squares (DMC) The variables of the development of the stock market and the growth The hypothesis the stock-exchange activity allows the explanation of the growth is (partially) supported. The stock market development effects on growth is low in least developed countries. It is, however, significant for the developed countries Levine and Zervos (1998) 42 countries Annual data between 1976 and 1993 MCO and GMM The variables of the banking sector and the growth of the real GDP per capita The banking sector development contributes positively to the growth of real GDP per capita Levine and Zervos (1998b) 47 countries Annual data between 1976 and 1993 MCO The levels of market liquidity and the variables of the banking sector, the growth of the real GDP per capita, the productivity and the physical capital stock The initial levels of market liquidity and banking sector development are positively and significantly correlated to the future growth of the real GDP per capita, the productivity and the physical capital stock. No strong impact about the size of the stock markets on the sources of growth was detected Table 1: Financial development and growth: A selective review of the main empirical research (Contd...) Aydi and Aguir: Financial Development and Economic Growth: The Empirical Evidence of the Southern Mediterranean Countries International Journal of Economics and Financial Issues | Vol 7 • Issue 3 • 2017 199 Authors Number of the country(s) Period Methodology Variables The empirical results Levine (1999) 49 countries Annual data between 1960 and 1989 GMM The variables of the financial intermediation and the economic growth Presence of a strong and significant positive correlation between the development of the financial intermediation explained and the growth Ram (1999) 95 countries Annual data between 1960 and 1989 MCO The variables; ratio of the liquid liabilities and the economic growth The correlation between the financial development (ratio of the liquid liabilities) and the growth is weakly negative or negligible McCaig and Stengos (2005) 71 countries Annual data between 1960 and 1995 GMM The ratio of the liquid liabilities, the credit to the private sector and the economic growth rate Positive effect of the finance on the growth when the financial development is measured by the ratio of the liquid liabilities or that of the credit to the private sector. The correlation between both variables is much lower when we describe the financial development by the asset ratio of commercial banks on the sum of this one with the asset of the central bank The panel data studies Beck et al. (2000b) 77 countries The quinquennial average data between 1960 and 1995 GMM on dynamic panel The variables of financial intermediation, the growth of productivity and the real GDP per capita, the capital accumulation and the savings Significantly positive and strong effect of the development of the financial intermediation on the growth of productivity and the real GDP per capita. Although less robust positive effect of this one on the capital accumulation and the growth of the savings Table 1: (continued...) (Contd...) Aydi and Aguir: Financial Development and Economic Growth: The Empirical Evidence of the Southern Mediterranean Countries International Journal of Economics and Financial Issues | Vol 7 • Issue 3 • 2017200 Authors Number of the country(s) Period Methodology Variables The empirical results Levine et al. (2000) 74 countries The quinquennial average data between 1960 and 1995 GMM on dynamic panel The variables of the financial intermediation and the growth of the real GDP per capita The existence of a correlation significantly positive between the development of the financial intermediation and the growth of the real GDP per capita Lopez and Spiegel (2002) 101 countries The quinquennial average data between 1965 and 1990 GMM on dynamic panel The variables of the financial development and the economic growth Significantly beneficial contribution of the financial development to the long-term growth. Short-term ambiguous relation Calderon and Liu (2003) 109 countries Data averaged over 5-10 years between 1960 and 1994 VAR models on panel, Geweke’s decomposition and Granger causality The variables of development of the financial intermediation and the economic growth Bidirectional causality between the development of the financial intermediation and the growth. Effect of the financial development on the growth stronger in developing countries, compared with the industrialized economies. The financial development affects the growth by acting essentially on the productivity growth Beck and Levine (2004) 40 countries The Quinquennial average data between 1976 and 1998 GMM on dynamic panel The variables of development of the financial intermediation, the stock markets liquidity and the economic growth The development of the financial intermediation and the stock markets liquidity allow the promotion of the growth Loayza and Ranciere (2004) 75 countries Annual data between 1960 and 2004 The PMG estimator (PMG) on dynamic panel The variables of the financial intermediation and the economic growth A relation significantly is positive in the long term, between the development of the financial intermediation and the growth, coexists with a short-term negative relationship in most countries in the sample Table 1: (continued...) (Contd...) Aydi and Aguir: Financial Development and Economic Growth: The Empirical Evidence of the Southern Mediterranean Countries International Journal of Economics and Financial Issues | Vol 7 • Issue 3 • 2017 201 Authors Number of the country(s) Period Methodology Variables The empirical results Stengos and Liang (2005) 66 countries The quinquennial average data between 1961 and 1995 Semi-parametric partially linear models The variables of financial development and economic growth Non-linear relationship between financial development and growth. Relation which depends on the financial development indicator used Saci et al. (2009) 30 developing countries The quinquennial average data between 1988 and 2001 GMM on dynamic panel The variables of the banking development, the economic growth and the market capitalization No effect or significantly negative of the banking development on the growth when we control the development of the stock market. Significantly positive effect of stock market development Hassan et al. (2011) Country with low or average income The quinquennial average data between 1980 and 2007 MCO, weighted least squares, VAR model Granger causality, FIR and variance decomposition The development of financial intermediation and economic growth Positive relationship between the development of the financial intermediation and the long-term growth AL-Malkawi and Abdullah (2011) 13 countries of the MENA region Annual data between 1985 and 2005 Pooled OLS, fixed effects model, random effects model The variables of development of the financial intermediation and the growth Positive relationship between the development of the financial intermediation and the growth Kar et al. (2011) 15 countries of the MENA region Annual data between 1980 and 2007 VECM model, MMG, Hurlin technique (2008) and approach of Kónya (2006) The variables of financial development and the economic growth The direction of causality between financial development and growth varies according to the financial development indicator used, as well as between the countries of the studied sample Time series studies Gupta (1984) 14 developing countries Quarterly data between 1961 and 1980 VAR model and Granger causality tests The variables of the finance and the economic growth The results show a causality which goes of the finance towards the growth. They support, in certain cases, the presence of reverse causality. A mutual causality is proved however rarely Table 1: (continued...) (Contd...) Aydi and Aguir: Financial Development and Economic Growth: The Empirical Evidence of the Southern Mediterranean Countries International Journal of Economics and Financial Issues | Vol 7 • Issue 3 • 2017202 3. DATA AND METHODOLOGY 3.1. Description of Data To examine the empirical connections between financial development and economic growth, we collect data for the real GDP per capita, the internal credit supplies in the private sectors, the market capitalization, M2/GDP and the inflation as the control variables, for a period from 1981 to 2014 with 34 observations. The variables of economic growth and the inflation are obtained from the World Bank database and the financial development variables of the database elaborated by Beck, Demirgüç-Kunt and Levine in 2013. The variable “GDP” design real GDP per capita. The GDP is the sum of gross value added generated by the productive sectors of an economy. It measures the efforts of economic output. Its relative variation from 1 year to another reflects the economic growth rate. This variable is the dependent variable of the model. The variable “CISP” indicates the credit to the private sector report to GDP the amount of the credit assigned to the private sector by banks and other nonbank financial institutions. Authors Number of the country(s) Period Methodology Variables The empirical results Jung (1986) 56 countries Annual data between 1950 and 1981 VAR model and Granger causality tests The variables of the finance and the economic growth Causality which goes of the financial development towards the growth in the least developed countries. A causality in the inverse sense for the developed countries Arestis and Demetriades and Luintel (1997) Germany and the United States Quarterly data between 1979 and 1991 VECM model and Johansen cointegration The financial development the real GDP Causality goes from financial development to real GDP for Germany, but in the opposite direction for the United States Arestis et al. (2001) 5 industrialized countries Quarterly data between 1972 and 1998 VECM model and Johansen cointegration The variables of the intermediation of banks, stock markets and economic growth The financial development of banks and stock markets promotes growth. It is the banks who contribute in a more significant and more important way to the growth process, compared with stock markets Thangavelu and Ang (2004) Australia Quarterly data between 1960 and 1999 VAR model and Granger causality The economic growth, the variables of the banking sector and the variables of development of the stock market The growth causes the development of the banking sector (according to Granger), while the development of the stock market causes the growth Ang and McKibbin (2007) Malaysia Annual data between 1960 and 2001 VECM, Johansen cointegration, Granger causality and ACP The economic growth and the development of the banking sector A long-term relation which goes from the growth to the development of the banking sector and not the opposite Source: Established by the authors from the literature review. VECM: Vector error correction model, VAR: Vector autoregression, OLS: Ordinary least squares, GMM: Generalized method of moment, GDP: Gross domestic product, PMG: Pooled mean group, FIR: Finite impulse response Table 1: (continued...) Aydi and Aguir: Financial Development and Economic Growth: The Empirical Evidence of the Southern Mediterranean Countries International Journal of Economics and Financial Issues | Vol 7 • Issue 3 • 2017 203 This ratio allows the level of activity of banking and non-banking financial intermediaries in the exercise of their function of channeling savings. It reflects the way in which domestic assets are distributed between the public and private sectors. It is based on the assumption that the more developed financial systems are those who attribute the most credits to private firms. The variable “CB” means the ratio of market capitalization (market cap): This ratio measures the size of the stock market. It is equal to the total value of parts quoted in stock exchange reported to the GDP. The use of this indicator supposes the existence of a positive correlation between the size of the stock market and its development. However, this is not always obvious. A wide stock market is not necessarily effective in the performance of his duties. It can, moreover, be developed strongly despite a small size (this one being explained by the presence of taxes preventing an adequate quotation in stock exchange, rather than a low efficiency of the market in the exercise of his functions). The variable “LIQ” indicates the liquid liabilities in the GDP: It is the ratio of the liquid liabilities of the economy in the GDP. This indicator takes into account the money supply (M2) and the liquid liabilities of financial institutions. The liquid liabilities is a measure of the financial depth or the global size of the financial system. The variable “Infl” indicates the Inflation rate. It is the variable which represents the macroeconomic politics. It is introduced into the model to get the impact of the macroeconomic stabilization on the economy. The inflation is a factor of worsening of the growth because it has a negative it has a negative impact on the actual value of the portfolio and the purchasing power of household incomes and thus on the growth. We use the consumer price index as the indicator which measures the inflation rate. 3.2. Methodology The main interest of this study is to analyze the impact of the financial development and the economic growth by using the models: VECM introduced by Johansen (1988) and the VAR model is proposed by Sims (1980). The advantage of the Johansen and Joselius cointegration procedure (1990) is that she allows on one hand testing the existence of one or several relations of cointegration between the various series. Secondly, the method of Johansen is a multivariate test which allows to determine the number of cointegration relationships between the selected series. Thus, this approach avoids the two step test applied in Engel-Granger procedure which allows to have a one cointegration relationship. This approach also has the advantage of taking into account the problem of simultaneity. Finally, the hypothesis of exogenous variables is not supported and it is not necessary to impose restrictions on the estimated coefficients to determine the short-term relationships. Let us consider a VECM model based on annual data for pib = (cisp, cb, liq, infl) given by: ∆ ∆y y B B yt t i t i t i p = + + +− −=∑αβ ε' 1 0 1 (1) Where Δ is the first difference of the operator, B0 is a column vector of 4 dimensions of determinist constant terms and Bi=i=1,…, P indicate matrices of order 4 of the short-term information parameters. αβ' is a matrix of order 4 of the long- term information parameters, Where α represent the speed of adjustment of the balance and β contains the long term or equilibrium coefficients. εt denotes a four-dimensional vector of residuals where εt~iid(0,Ω). The rank (αβ) = r is the number of cointegrating vectors which can vary according to the country and to the nature of the variable tested. If r = 0, the time-series variables are not cointegrated, in this case, the variables must first be differentiated and we have the VAR in the difference. In the first stage, we use the traditional unit root testing of augmented Dickey-Fuller (ADF), Phillips-Perron (PP) and Kwiatkowski-Phillips-Schmidt-Shin (KPSS) tests to verify the stationarity of all the variables. Secondly, we apply other similar tests as endogenous break unit root tests of Lee and Strazicich breaks (2003; 2004) to avoid “spurious emissions” of conventional unit root tests. We proceed in the second stage to determine the length of delay of the VAR of the VECM models using the information criterion Schwartez (CIS), for variables growth rate of GDP per capita, internal credits supplies in the private sector, market capitalization, liquid liabilities and inflation rate contains a unit root. Then we apply the Johansen cointegration test to determine the number of cointegrating vectors (rang [αβ'] = r) using two different statistics: The trace statistic and the maximal eigenvalue statistic. In the third step, the estimated VECM between variables real GDP per capita, the domestic credit to private sector, market capitalization, examining the impulse response functions (IRF) obtained by estimating the previous VECM. 4. EMPIRICAL RESULTS 4.1. Results of the Unit Root Tests The ADF unit root tests, PP and KPSS variable on each level and in first difference has been made, and this for all countries in the sample. The test results are reported in Table 2. The statistics of ADF, PP and KPSS suggest that all variables are integrated of order 1, I(1). Except for the cases of Egypt; liquid liabilities and variable inflation rates which are stationary in level I(0) at the 5% threshold and 10% respectively. The variable inflation rate seems to be integrated I(0) to the same tests in the case of Jordan and Lebanon. The history of the series, liquid Liabilities and inflation rate in time, show that for every country the series are not really fixed trend in the level. The Figure 1 indicates a presence of ruptures in all the series of variable liquid liabilities and inflation rate. The distribution of the series of variables presented by the Figure 1 confirms the non-stationarity. It indicates in fact, the existence of a trend for the majority of the series. It also indicates the high probability existence of one or several structural rupture. This incites us to make the test of Lee and Strazicich (2003). Which Aydi and Aguir: Financial Development and Economic Growth: The Empirical Evidence of the Southern Mediterranean Countries International Journal of Economics and Financial Issues | Vol 7 • Issue 3 • 2017204 will allow us to test the stationarity in the presence of structural rupture. The conventional unit root tests (ADF, PP and KPSS) are not able to reject the null hypothesis when the structural repture are present. These tests conduct their critical values are assuming no rupture under the null hypothesis. Consequently, in the presence of a unit root with rupture, they tend to reject the null hypothesis suggesting that the time series is stationary around trend when it is non-stationary with a rupture. For this reason, we conduct tests for endogenous rupture in unit root. Christiano (1992), Perron and Vogelsang (1992), Zivot and Andrews (1992) have developed methods to determine a repture point and to test the presence of a unitarian root when the process has a constant broken or trendy and demonstrated that their tests are robust and efficient than the conventional unit root tests. To avoid this problem and to examine the potential presence of rupture, we use in this paper the LM unit root test with two breaks endogenous proposed by Lee and Strazicich (2003; 2004). This result seems to be affected by ruptures under the null hypothesis. We find that significant structural ruptures provided for both series: Liquid liabilities and inflation rate. Concerning unit root tests ADF, PP, KPSS and LM, the results conclude in favor of the unit root I(1) for all series in all countries. 4.2. Johansen Cointegration Test The study of the cointegration allows to test the existence of a long- term stable relationship between the variables integrated of order 1 I(1). There are several tests of cointegration, the most general being that of Johansen. Whatever the chosen test, it has meaning only on stationary series in first difference. Consequently, the analysis of the cointegration allows to identify the true relationship between variables, by searching the existence of a cointegration vector and by eliminating its effect if necessary. Two series x and y are called cointégrées if following both conditions are verified: They are affected by a stochastic trend of the same order of integration and a linear combination of these series and a linear combination of these series can be reduced to a series of order of integration lower. Finally, the Johansen cointegration test uses two statistics: Statistics trace and the maximum eigenvalue whose order is d’exterminé by the Schwarz criterion (SC). The unit root tests ADF, PP, KPSS and LM (Lee and Strazicich) prove that all variables contain a unit root, then we test cointegration in each VECM using both the trace and the maximum eigenvalue. Results of the application of Juselius (1990) and Johansen approach are presented in Table 3. The Table 3 includes the ranks given in the first line, the number of cointegrating vectors in line 2, eigenvalue and track statistics for each selected country. The critical value is mentioned using asterisks. The null hypothesis is that the number of cointegrating relationship is equal to r, which is given in the “maximum rank” observed in the first row of Table 3. The alternative is that there are more cointegration relationships r. We reject the null hypothesis if the trace statistic is greater than the critical value. We begin by testing H0: r 0. If the null hypothesis is rejected, we repeat for H0: r = 1. The process continues for T ab le 2 : S ta ti on ar it y te st A D F, P P an d K P SS (v ar ia bl es : G D P, C IS P, L IQ , C B a nd I N F ) C ou nt ry R ea l G D P pe r ca pi ta D om es ti c cr ed it to p ri va te se ct or M ar ke t c ap it al iz at io n L iq ui d lia bi lit ie s to G D P In fla ti on r at e A D F P P K P SS A D F P P K P SS A D F P P K P SS A D F P P K P SS A D F P P K P SS V ar ia bl es in le ve ls E gy pt −0 .7 14 −1 .3 83 0. 70 0* * −1 .7 35 −2 .0 56 0. 37 2* −1 .5 95 −1 .6 37 0. 48 4* * −3 .5 22 ** −3 .4 88 ** 0. 28 3 −2 .7 29 * −2 .6 56 * 0. 40 5* Is ra el −0 .4 36 −0 .3 49 0. 69 5* * −1 .0 64 −0 .9 86 0. 52 7* * −2 .3 75 −2 .3 75 0. 55 3* * −0 .4 83 −0 .5 10 0. 63 4* * −1 .7 65 −1 .9 96 0. 46 2* * Jo rd an −1 .2 83 −0 .4 2 0. 37 2* * −2 .2 88 −2 .0 39 0. 67 2* * −2 .0 19 −2 .1 01 0. 40 3* −2 .6 87 * −1 .7 13 0. 53 5* * −3 .6 85 ** * −3 .4 75 ** 0. 24 4 L eb an on −2 .2 86 −2 .3 35 0. 62 2* * −0 .9 87 −1 .0 98 0. 58 9* ** −2 .5 93 −2 .0 52 0. 16 −0 .9 05 −0 .7 60 0. 57 9* * −3 .3 46 ** −3 .3 35 ** 0. 22 7 M or oc co 1. 43 5 1. 3 34 0. 65 3* * 0. 01 3 0. 05 1 0. 64 7* * −1 .1 78 −1 .2 75 0. 64 5* * −0 .6 46 −0 .6 03 0. 65 8* * −2 .3 87 −2 .1 46 0. 66 3* * Tu ni si a 1. 07 1. 15 9 0. 65 7* * −1 .2 55 −2 .5 49 0. 35 9* −2 .3 48 −1 .4 68 0. 51 2* * 0. 57 3 −0 .2 31 0. 64 1* * −1 .3 15 −1 .7 20 0. 44 8* Tu rk ey −0 .4 16 −0 .1 94 0. 68 ** 4. 03 9 4. 08 1 0. 49 ** −0 .5 51 −2 .6 14 0. 73 0* ** 0. 25 8 0. 40 4 0. 68 8* * −2 .2 05 −2 .3 75 0. 39 9* V ar ia bl es in fi rs t di ff er en ce s E gy pt −4 .6 78 ** * −4 .6 88 ** * 0. 15 6 −3 .9 46 ** * −0 .3 96 7* ** 0. 29 1 −5 .4 18 ** * −5 .4 18 ** * 0. 11 5 −5 .8 32 ** * −5 .8 45 ** * 0. 27 6 −9 .5 09 ** * −9 .5 09 ** * 0. 22 2 Is ra el −5 .1 36 ** * −6 .2 06 ** * 0. 15 5 −6 .4 25 ** * −6 .8 28 ** * 0. 37 3* −6 .2 53 ** * −6 .4 89 ** * 0. 11 0 −5 .1 15 ** * −5 .1 11 ** * 0. 10 4 −4 .8 68 ** * −8 .1 86 ** * 0. 5* * Jo rd an −3 .4 46 *( b) −4 .3 28 ** * 0. 28 −5 .1 28 ** * −3 .7 34 ** * 0. 32 0 −6 .7 79 ** * −6 .7 45 ** * 0. 09 4 −2 .8 73 * −4 .8 88 ** * 0. 13 0 −6 .7 28 ** * −1 5. 36 7* ** 0. 5* * L eb an on −4 .8 02 ** * −4 .7 63 ** * 0. 08 6 −5 .6 53 ** * −5 .6 53 ** * 0. 18 6 −3 .2 69 ** −3 .2 69 ** 0. 07 −4 .3 05 ** * −4 .2 03 ** * 0. 17 4 −8 .6 38 ** * −1 0. 88 9* ** 0. 17 5 M or oc co −1 1 .4 28 ** * −1 0. 44 1* ** 0. 32 −6 .1 8* ** −6 .1 65 ** * 0. 11 9 −4 .9 02 ** * −4 .8 67 ** * 0. 06 7 −6 .7 30 ** * −6 .7 11 ** * 0. 04 2 −4 .5 69 ** * −1 3. 91 8* ** 0 .3 76 * Tu ni si a −6 .6 89 ** * −5 .6 93 ** * 0. 37 * −1 .1 09 −5 .2 35 ** * 0. 08 8 −4 .2 36 ** * −6 .1 53 ** * 0. 29 6 −4 .2 63 ** * −5 .5 73 ** * 0. 18 1 −9 .1 69 ** * −9 .7 98 ** * 0. 5* Tu rk ey −6 .3 64 ** * −7 .8 13 ** * 0. 11 9 −3 .6 97 ** (b ) −3 .6 57 ** (b ) 0. 45 5* −7 .0 12 ** * −1 5. 80 4* ** 0. 30 8 −4 .4 04 ** * −4 .2 66 ** * 0. 18 8 −7 .7 91 ** * −8 .3 07 ** * 0. 10 7 A D F: A ug m en te d D ic ke y- Fu lle r, PP : P hi lli ps -P er ro n, K PS S: K w ia tk ow sk i- Ph ill ip s- Sc hm id t- Sh in , G D P: G ro ss d om es tic p ro du ct . * ,* * an d ** * In di ca te a n ac ce pt an ce o f t he h yp ot he si s at th e th re sh ol d of 1 0% , 5 % a nd 1 % re sp ec tiv el y. (b ) d en ot es th e st at io na ry w ith T re nd a nd in te rc ep t. th e le ng th o f d el ay in a ll th e te st s w as s el ec te d ac co rd in g to th e in fo rm at io n cr ite ri on S ch w ar te z Aydi and Aguir: Financial Development and Economic Growth: The Empirical Evidence of the Southern Mediterranean Countries International Journal of Economics and Financial Issues | Vol 7 • Issue 3 • 2017 205 Figure 1: Evolution of the series of variable by country. (a) Variable interest: Liquid liabilities (LIQ). (b) Variable interest: Inflation rate (INF) b a r = r = 2... 3 = 4 and r. The process ends when a test is not rejected. Existing of one or several cointegration vectors explains that the variables have a long-term relationship and we should continue to use VECM. Aydi and Aguir: Financial Development and Economic Growth: The Empirical Evidence of the Southern Mediterranean Countries International Journal of Economics and Financial Issues | Vol 7 • Issue 3 • 2017206 The results of cointegration show that there are at least two cointegration vectors with an interception and/or trend in all the countries. Therefore, we can conclude that there are at least two cointegrating vectors for all the selected countries. Based on the results of Johansen cointegration tests, we conclude that the VCEM can be applied to all countries specifically for even the answers the impulse responses of the domestic credit provides private and market capitalization on economic growth sectors (Table 4). Table 3: The unit root test with two breaks of Lee and Strazicich Series One-break Two breaks Model A Model B Model A Model B t-stat break t-stat break t-stat break t-stat break Liquid liabilities to GDP Egypt −2.230 2004 −4.385 2005* −2.472 2004 2007 −5.674*** 1999 2010 Israel −3.576** 2009 −17.25*** 2009 −3.587** 1995 2009 −17.911*** 1984 2009 Jordan −4.239*** 1985 −6.082*** 1985 4.039** 1985 1996 −6.790*** 1984 1989 Lebanon −3.993** 1989 −4.302* 2004 −4.224* 1988 2005 −5.852*** 1997 2007 Morocco −4.249*** 2006 −4.809** 2000 −4.927*** 1997 2006 −6.421*** 1989 2004 Tunisia −3.257* 1985 −2.750 1993 −3.384* 1985 2008 −17.114*** 1986 2010 Turkey −1.855 1998 −5.812*** 2010 −1.994 1992 1998 −8.865*** 2004 2010 Inflation rate Egypt −4.103** 2005 −4.36 0* 2006 −5.329*** 1997 2007 −7.157*** 1995 2008 Israel −5.190*** 2005 −5.014** 1990 −5.787*** 1987 2005 −7.396*** 2000 2009 Jordan −3.367* 2010 −4. 188* 1987 −3.730** 1985 2010 −4.375* 1987 2001 Lebanon −6.679*** 1991 −6.880*** 1992 −6.932*** 1992 1997 −8.530*** 1995 2000 Morocco −4.847*** 1994 −4.887** 1991 −5.126*** 1994 2000 −5.916*** 1996 2002 Tunisia −3.283* 1991 −5.015*** 1987 −4.188** 1988 1998 −6.053*** 1987 1998 Turkey −4.471*** 2006 −7.282*** 2004 −6.644*** 2006 2009 −10.335*** 2004 2009 Model A: Change in the interception. Model C: Change in the constant and the trend. The critical values for the unit root test LS with a break are indicated in Lee and Strazicich (2004, Table 1). The critical values for the unit root test LS with two breaks, which appear in Lee and Strazicich (2003, Table 2), depend on the location of the rupture. *,**,*** indicate the level of signification respectively at 10%, 5% and 1%. GDP: Gross domestic product Table 4: Results of Johansen cointegration test (variables: GDP, CISP, LIQ, CB and INF) Country r=0 r≤1 r≤2 r≤3 r≤4 1 2 1 2 1 2 1 2 1 2 Egypt Trace statistic 80.23644*** 106.5183*** 41.94786 60.36584* 20.04274 35.90485 6.920876 17.42934 1.515200 4.320857 Max-Eigen stat 38.28858** 46.15243*** 21.90512 24.46099 13.12186 18.47552 5.405675 13.10848 1.515200 4.320857 Israel Trace statistic 76.00426** 85.80266* 47.56137* 53.52832 20.79332 25.90573 5.938574 10.83198 0.333081 4.067754 Max-Eigen stat 28.44289 32.27434 26.76804* 27.62259 14.85475 15.07375 5.605492 6.764229 0.333081 4.067754 Jordan Trace statistic 95.67284*** 117.1330*** 56.47780*** 65.87977** 31.77948** 40.05070* 12.25967 20.46788 3.371765* 8.140777 Max-Eigen stat 39.19505* 51.25328*** 24.69832 25.82907 19.51981* 19.58282 8.887900 12.32711 3.371765* 8.140777 Lebanon Trace statistic 91.68223*** 130.8359*** 45.18321* 77.03425*** 21.06532 37.11210 7.825086 16.17432 0.310289 5.807031 Max-Eigen stat 46.49901*** 53.80161*** 24.11789 39.92215*** 13.24024 20.93778 7.514796 10.36729 0.310289 5.807031 Morocco Trace statistic 79.29406*** 103.5950*** 36.13360 53.67280 18.38017 28.59204 7.466910 16.18721 0.044629 5.616787 Max-Eigen stat 43.16046*** 49.92217*** 17.75343 25.08075 10.91326 12.40483 7.422281 10.57042 0.044629 5.616787 Tunisia Trace statistic 88.95627*** 101.9829*** 51.15970** 60.64219* 19.29113 26.31669 5.585769 12.52694 0.501403 2.813502 Max-Eigen stat 37.79658** 41.34072** 31.86856** 34.32550** 13.70536 13.78975 5.084366 9.713433 0.501403 2.813502 Turkey Trace statistic 84.34933*** 111.6427*** 47.10472* 74.25474*** 21.02951 39.46615 7.886801 19.48983 0.029780 7.612779 Max-Eigen stat 37.24461** 37.38793* 26.07521* 34.78860** 13.14270 19.97632 7.857021 11.87705 0.029780 7.612779 1: Model with one interception, 2: Model with an interception and a linear trend, r: Number of cointegrating vector. ** and *** indicate the reject of the null hypothesis at the threshold of 10%, 5% and 1% respectively. The length of delay in all the tests was selected according to the information criterion Schwartez. GDP: Gross domestic product Aydi and Aguir: Financial Development and Economic Growth: The Empirical Evidence of the Southern Mediterranean Countries International Journal of Economics and Financial Issues | Vol 7 • Issue 3 • 2017 207 Figure 2: Domestic credit shock provided to the private sector and market capitalization Aydi and Aguir: Financial Development and Economic Growth: The Empirical Evidence of the Southern Mediterranean Countries International Journal of Economics and Financial Issues | Vol 7 • Issue 3 • 2017208 5. THE EFFECT OF THE DOMESTIC CREDIT SHOCKS PROVIDED TO THE PRIVATE SECTOR AND MARKET CAPITALIZATION ON ECONOMIC GROWTH To evaluate the effect of the domestic credit shocks provided to the private sector and market capitalization on economic growth for Egypt, Israel, Jordan, Lebanon, Morocco, Tunisia and Turkey. We use IRF and analyze the impact of these shocks on the economic growth of countries in the region. The two columns of Figure 2 respectively describe the impact of the domestic credit shocks provided to the private sector and market capitalization. 5.1. Domestic Credit Shock (CISP) Figure 2 shows that the domestic credit shock (CISP) (first column) have a significant negative effect on economic growth in Egypt, Israel, when they have a positive effect only to Jordan. The effect of the shocks on the economic growth for Morocco and Tunisia, is however not significant. In fact the shape of the impulse response curve is volatile close to zero, sometimes positive and sometimes negative. For the case of Lebanon and Turkey the impact is mixed, what is positive in first period become negative for the second period for stabilizes in the sixth period in Lebanon is still negative in Turkey. 5.2. Market Capitalization Shock (CB) The shocks on market capitalization (the second column) has a significantly positive effect on the economic growth for Egypt, Israel, Lebanon and Morocco, the shock is triggered in the first period to reach a maximum value in the second period and creates a peak, to gradually return to its equilibrium position from the third period. The effect of shocks on the economic growth for Jordan and Tunisia, is however not significant. Indeed, the shape of the impulse response curve generally remains close to the abscissa axis of value zero order. For Turkey, the triggering of the impulse response in half of the second period, has reached a significantly negative maximum value for created a peak in the third period and return to its equilibrium the end of the period. The main results for the impulse responses show that a shock on domestic credit provides private sectors and market capitalization influences positively or negatively the economic growth has short term, the answers are clearer for credits that essentially affects the economies of the region, as do not incur a great tradition of the stock change. 6. CONCLUSION The released results of our estimate shows that there is a long- term relationship between financial development and economic growth. Indeed functions of impulse responses shows that a shock on the domestic credit variable relating to the private sector has a significant effect on economic growth that a shock on the stock capitalization, but overall, the financial sector in the countries of the region PSM continues to play a less important role than in other economies with similar income levels, but we notice considerable écarts between the countries of the region in terms of financial sector development level. The financial systems in the region remain dominated by banks, and financial intermediation is still in the development stage according to the international standards. However, the banking sector does not occupy an important place in the stimulation of economic development. In spite of the privatizations, the participation of States in bank’s capital continues to be higher than in the other countries of similar level. The countries of the region PSM have to develop strategies to promote innovation, competition and the expansion of coverage of the financial sector. By considering the particular situation of each of the countries, such strategies could include opening bank markets to foreign and local new entrants and promoting better credit culture of the credit to facilitate the access to the finance, associated with more effective prudential supervision. In most of the countries, all this must be completed by legal and institutional reforms in the domains of the accounting, auditing, financial probity and corporate governance, in order to promote transparency and accountability. Stock markets in the region are relatively new, and the market capitalization, the value of exchanges and companies quoted remain low compared with high income countries. Globally issuing shares and bonds is still a little used method of fund-raising in the region, leaving the banking sector without competition. Concerning policy implications, we must draw proposals based on the results. It is clear that improving the performance of the financial system in the region is absolutely essential in order to allow financial development as a growth stimulant. Therefore, PSM need to improve the credit allocation process through the privatization of domestic banks, by strengthening credit regulation and by increasing competition in the banking sector. In addition, a prerequisite appears to be that the regulatory infrastructure is well developed and that measures are being taken to reduce the extreme volatility of stock prices in order to allow the stock market in the PSM regions to stimulate economic growth. REFERENCES Abu-Bader, S., Abu-Qarn, A. (2006), Financial Development and Economic Growth Nexus: Time Series Evidence from Middle Eastern and North African Countries. MPRA Paper N_972. Acaravci, S., Ozturk, I., Acaravci, A. (2009), Financial development and economic growth: Literature survey and empirical evidence from sub-Saharan African countries. South African Journal of Economic and Management Sciences, 12(1), 11-27. Al-Malkawi, H.A.N., Abdullah, N. (2011), Finance-growth nexus: Evidence from a panel of MENA countries. International Research Journal of Finance and Economics, 63, 129-139. Ang, J.B., McKibbin, W.J. (2007), Financial liberalisation, financial sector development and growth: Evidence from Malaysia. Journal of Development Economics, 84(1), 215-233. Arestis, P., Demetriades, P., Luintel, K. (2001), Financial development and economic growth: The role of stock markets. Journal of Money, Credit, and Banking, 33, 16-41. Aydi and Aguir: Financial Development and Economic Growth: The Empirical Evidence of the Southern Mediterranean Countries International Journal of Economics and Financial Issues | Vol 7 • Issue 3 • 2017 209 Arestis, P., Demetriades, P. (1997), Financial development and economic growth: Assessing the evidence. Economic Journal, 107(442), 783-799. Atje, R., Jovanovic, B. (1993), Stock markets and development. European Economic Review, 37(2-3), 632-640. Beck, T., Levine, R., Loayza, N. (2000b), Finance and the sources of growth. Journal of Financial Economics, 58(1-2), 261-300. Beck, T., Levine, R. (2004), Legal Institutions and Financial Development in Claude. Calderon, C., Liu, L. (2003), The Direction of Causality between financial development and economic growth. Journal of Development Economics, 72, 321-334. Christiano, LJ. (1992), Searching for breaks in GNP. Journal of Business and Economics Statistics, 10, 237- 250. Demetriades, P.O., Hussein, K. (1996), Does financial development cause economic growth? Time-series evidence from sixteen countries. Journal of Development Economics, 51(2), 387-411. Goldsmith, R. (1969), Financial Structure and Development. New Haven, CT: Yale University Press. Granger, C.W.J., Lin, J. (1995), Causality in the long run. Econometric Theory, 11, 530-536. Gupta, K.L. (1984), Finance and Economic Growth in Developing Countries. London: Croom Helm. Harris, R. (1997), Stock market and development: A re-assessment. European Economic Review, 41, 139-146. Hassan, M.K., Sanchez, B., Yu, J. (2011), Financial development and economic growth: New evidence from panel data. The Quarterly Review of Economics and Finance, 51(2011), 88-104. Available from: http://www.elsevier.com/locate/qref. Johansen, S. (1988), Statistical and hypothesis testing of cointegration vectors. Journal of Economic Dynamics and Control, 12, 231-254. Johansen, S., Juselius, K. (1990), Maximum likelihood estimation and inference on cointegration – with applications to the demand for money. Oxford Bulletin of Economics and Statistics, 52(2), 169-210. Jung, W.S. (1986), Financial development and economic growth: International evidence. Economic Development and Cultural Change, 34, 336-346. Kar, M., Nazlioglu, S., Agir, H. (2011), Financial development and economic growth nexus in the MENA countries: Bootstrap panel granger causality analysis. Economic Modelling, 28(1-2), 685-693. Lee, J., Strazicich, M.C. (2003), Minimum LM unit root test with two structural breaks. Review of Economics and Statistics, 63, 1082- 1089. Lee, J., Strazicich, M.C. (2004), Minimum LM Unit Root Test with One Structural Break. Working Paper, Department of Economics, Appalachain State University. 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, E.S. (1998b), Capital control liberalization and stock development. World Development, 26, 84-1169. Levine, R. (1999), Law, finance, and economic growth. Journal of Financial Intermediation, 8(1-2), 8-35. Levine, R., Zervos, E.S. (1998), Stock markets, banks and economic growth. American Economic Review, 88(3), 578-580. Loayza, N., Ranciere, R. (2004), Financial Development, Financial Fragility and Growth. Paper Provided by World Bank in its series Policy Research Working Paper Series with Number 3431. Lopez, J.A., Spiegel, M.M. (2002), Analyse Cross-Section Data and Find that Financial Development does Mitigate Economic Fluctuations. Luintel, K.B., Khan, M. (1999), A quantitative re-assessment of the finance-growth nexus: Evidence from a multivariate VAR. Journal of Development Economics, 60(2), 381-405. McCaig, B., Stengos, T. (2005), Financial intermediation and growth: Some robustness results. Economics Letters, 88, 306-312. Mckinnon, R.I. (1973), Money and Capital in Economic Development. Washington, DC: The Brookings Institution. Neusser, K., Kugler, M. (1998), Manufacturing growth and financial development: Evidence from OECD countries. Review of Economics and Statistics, 80, 638-646. Ozturk, I. (2008), Financial development and economic growth: Empirical evidence from Turkey. Applied Econometrics and International Development, 8(1), 85-98. Perron, P., Vogelsang, T.J. (1992), Nonstationarity and level shifts with an application to purchasing power parity. Journal of Business and Economic Statistics, 10, 301-320. Ram, R. (1999), Financial development and economic growth. The Journal of Development Studies, 35(4), 164-174. Rousseau, P.L., Wachtel, P. (1998), Financial intermediation and economic performance: Historical evidence from five industrialized countries. Journal of Money, Credit and Banking, 30(4), 657-678. Saci, K., Giorgioni, G., Holden, K. (2009), Does financial development affect growth? Applied Economics, 41(13), 1701-1707. Shaw, E. (1973), Financial deepening in economic development. New York: Oxford University Press. Sims, C. (1980), Macroeconomics and reality. Econometrica, 48, 1-48. Stengos, T., Liang, Z. (2005), Financial intermediation and economic growth: A semi parametric approach. New Trends in Macroeconomics. Berlin, Germany: Springer Heidelberg. p39-52. Thangavelu, S.M., Ang, B.J. (2004), Financial development and economic growth in Australia: An empirical analysis. Empirical Economics, 29(2), 247-260. Xu, Z. (2000), Financial development, investment, and economic growth. Economic Inquiry, 38(2), 331-344. Zivot, E., Andrews, D. (1992), Further evidence on the great crash, the oil-price shock and the unit root hypothesis. Journal of Business and Economic Statistics, 10, 251-270.