Do Remittances Matter for Financial Development in the MENA Region? Panel Cointegration and Causality Analysis

Transcription

1 The Empirical Economics Letters, 0(5): (May 20) ISSN Do Remtances Matter for Financial Development in the MENA Region? Panel Cointegration and Causaly Analysis Huseyin Agir Department of Economics, K.Maras Sutcu Imam Universy K.Maras, Turey Muhsin Kar Department of Economics, Cuurova Universy, Adana, Turey Saban Nazlioglu Department of Econometrics, Pamuale Universy, Denizli, Turey Abstract: This paper investigates the impacts of remtances on financial development in the MENA region by utilizing the panel cointegration and causaly methods and finds out that while the remtances and economic growth cause financial development in both the short- and long-run, financial development and economic growth induce the worer s remtances to the home country in the long-run. Keywords: Financial Development, Remtances, Panel Cointegration, Causaly, MENA JEL Classification Number: F24, F4, E44, C33. Introduction Volume of remtance flows by migrant worers from the host countries into the home countries has rapidly risen over the years (Brown, 2006; World Ban 2006) and this sort of private transfers has became one of the important resources for economic development (Chami et al., 2008). Recently, both theoretical and empirical leratures on this subject have mainly concentrated on whether macroeconomic variables and condions in both the home and the host country determine magnude of remtances (El-Saa and McNabb, 999; Sayan, 2004; Vargas-Silva and Huang, 2006). In addion, the impacts of remtances on the receiving economies are investigated through three channels. Remtances might alleviate poverty by increasing the recipient family s income and living standards (Adams and Page, 2005; Gupta et.al., 2009, Rao and Hassan, 200), increase the growth rate by

2 The Empirical Economics Letters, 0(5): (May 20) 450 reducing the volatily (Hnatovsa and Loayza, 2003; Chami et al., 2008), and negatively affect the growth rate by appreciating the exchange rate (Amuedo-Dorantes and Pozo, 2004; Lopez et.al., 2007; Lartey et. al., 2008). The lerature also argues that remtances may induce financial development through a few channels (Aggarwal et al., 2006; Giuliano and Ruiz Arranz, 2009; Gupta et.al., 2009; Acosta et.al., 2009). First of all, aggregate level of deposs may increase if remtances are held in the bans. Second, this rise in the deposs may enhance an increase in the total amount of the cred channeled into the private sector in the receiving country. This implies that a well-functioning financial maret may play an important role to direct the remtances to the projects that yield the highest return and therefore enhance growth rates. Third, remtances may introduce new households wh the baning system and, in turn, this familiary may result in both more depos and new cred opportunies in the sector. The aim of this paper is to examine to what extend the remtances contribute to the financial development in the MENA (Middle East and North Africa) region which receives a lion share from the regional distribution of the world-wide remtances, mainly from the European Union. To the best of our nowledge, this is the first empirical study which investigates whether the remtances contribute to financial development in the MENA region and how the direction of causaly, if any, between the remtances and financial development is. 2. Model and Data In order to empirically investigate impact of the remtances on financial development, this paper augments the tradional finance-growth model (Patric, 966; Trew, 2006) by adding the remtances. Thereby, following model can be wrten to examine the sensivy of financial development to the remtances: ln FD = α + ε () 0i + αi lny + α 2i ln REM ( i =,2,3,...,N ) ; ( t =,2,3,...,T ) where FD, Y, and REM stand for financial development, income level and remtances, respectively. The annual data covering the period for nine MENA countries (Algeria, Egypt, Israel, Jordan, Morocco, Sudan, Syria, Tunisia, and Turey), restricted by the availabily of the number of countries and the time period, is compiled from the World Development Indicators. Real per capa income and remtances are used as proxies for the economic growth and the remtances.

3 The Empirical Economics Letters, 0(5): (May 20) 45 Before going into the empirical analysis, one should highlight that financial development is a multifaceted issue and has not a direct measurement at hand. Instead, a number of proxies to catch up various dimensions of financial development have been widely used. This paper develops a Financial Development Index (FD) by utilizing the principle component analysis. The index is constructed from the widely employed six indicators of financial development: ) The ratio of narrow money to income, 2) the ratio of quasi money to income, 3) the ratio of broad money (M2) to income, 4) the ratio of depos money ban liabilies to income, 5) the ratio of private sector cred to income, and 6) the ratio of domestic cred to income. The financial sector is mainly dominated by the baning sector in the MENA countries and therefore these indicators may well capture the developments in the financial sector of these countries. The data on the financial variables are compiled from the IMF s International Financial Statistics online database. 3. Methods and Findings This study utilizes panel un root, cointegration, and causaly tests since panel data sets increase the statistical power of tests by combining information from both cross-section and time dimensions. The empirical analysis is carried out in three steps: Un root analysis, cointegration analysis, and causaly analysis. 3.. Panel Un Root Analysis The stationary properties of the variables are examined by the panel un root tests of Levin et al. (2002) (LLC) and Im et al. (2003) (IPS). Results in table show that the series are I() which signals a possible cointegration relation among the variables. Table : Panel Un Root Tests LLC IPS lnfd [0.005] -0.5 [0.4368] lny 2.60 [0.9954] 5.30 [.0000] lnrem [0.0] 0.70 [0.7588] lnfd -2.4 [0.0000] [0.0000] lny -.2 [0.0000] -.28 [0.0000] lnrem [0.0000] [0.0000] Note: Numbers in bracets are p-values. Schwarz Bayesian Crerion was used to determine the optimal lag lengths.

4 The Empirical Economics Letters, 0(5): (May 20) Panel Cointegration Analysis To test for the null of no-cointegration in the panel, the cointegration tests of Pedroni (999) are employed. The cointegration statistics except group p-test in table 2 provide evidence on the steady-state equilibrium in the long-run among financial development, the economic growth, and the remtances. Table 2: Panel Cointegration Tests Whin-dimension tests Constant Constant and Trend Panel-ν *** Panel- ρ -.4 * -.04 Panel-PP *** *** Panel-ADF -2.3 ** -3.0 *** Between-dimension tests Group- ρ Group -PP *** ** Group-ADF -2.3 ** *** Note: ***, ** and * indicates the statistical significance at, 5, and 0% levels respectively. The cointegration parameters are estimated by the group-mean panel FMOLS and DOLS methods developed by Pedroni (2000 and 200). The panel FMOLS estimator can be constructed as ˆ * = N * β GFM N β FMi where β * i= FMi is obtained from the time series FMOLS estimation of the equation () for each country. The panel DOLS estimator requires estimating the following model by the OLS for each country. 0i i 2i K ii ϖi = Kii K ii λi = Kii lnfd = β + β lny + β lnrem + lny + lnrem + ε * (2) where K i and K i are leads and lags. The panel DOLS estimator can be constructed as ˆ * = N * * β GD N β Di where β Di is obtained from the estimation of equation (2). i= The cointegration parameters in table 3 indicate that financial development is posively associated wh the economic growth and the remtances. More specifically, the panel DOLS (FMOLS) estimator shows that percent increase in the income and the remtances respectively stimulates financial development by 3.5 (3.24) and 0.78 (0.40) percent. The

5 The Empirical Economics Letters, 0(5): (May 20) 453 findings also show that the impact of income growth on financial development is greater than that of the remtances in the MENA countries. Table 3: Panel Cointegration Estimation lny lnrem Panel DOLS 3.5 *** 0.78 *** Panel FMOLS 3.24 *** 0.40 *** Note: Leads and lags were set to for the panel DOLS estimator. *** denotes statistical significance at percent level Panel Causaly Analysis The cointegration relationship implies that there are causal interactions among the variables. If the variables are cointegrated, one needs to estimate a vector error correction model (VECM) by augmenting the VAR model wh one-lagged error correction term. The panel VECM can be wrten as follows: = δi + δip lnfd p + δ2ip lny p + δ3ip lnrem p + ϕi ˆ ε lnfd + υ = δ2i + δ2ip lny p + δ22ip lnfd p + δ23ip lnrem p + ϕ2i ˆ ε lny + υ = δ3i + δ3ip lnrem p + δ32ip lnfd p + δ33ip lny p + ϕ3i ˆ ε lnrem + υ where is the optimal lag length(s) and εˆ is the residuals from the panel FMOLS estimation of the equation (). This specification allows identifying for both the short- and long-run causalies. The short-run causaly from one variable to other variable is tested wh a Wald test by imposing zero restriction on the parameters of the first-differenced variables. The long-run causaly is examined by statistical significance of the t-statistic of the error correction coefficients (ϕ ) (ECT). The results for causaly analysis in Table 4 indicate the short-run uni-directional causaly from the economic growth and from the remtances to financial development that lasts into the long-run. The findings fail to show any causal feedbac from the financial development to neher the economic growth nor the remtances in the short-run. Even though remtances are not sensive to financial development and the economic growth in the short-run, financial development and the economic growth are cause of remtances in the long-run. 2 3 (3) (4) (5)

6 The Empirical Economics Letters, 0(5): (May 20) 454 Table 4: Panel Granger Causaly Short-run causaly Long-run causaly lnfd lny lnrem ECT lnfd [0.0000] 6.06 [0.038] 0.08 ** lny 2.36 [0.242] 0.0 [0.986] lnrem 0.6 [0.6822] 0.52 [0.4670] 0.4 *** Note: The p-values are in bracets and parentheses. *** and ** indicate the statistical significance at and 5 percent levels respectively. 4. Conclusion This paper empirically investigates the relationships among the remtances, financial development and the economic growth for nine MENA countries to see whether and how the remtances might affect the financial development. The empirical results obtained from panel cointegration and panel causaly tests for the period show that (i) the demand following hypothesis of Patric (966:74), stated that as economy grows generates addional and dew demand for financial services, which bring about a supply response in the growth of the financial system, is supported in the MENA region and (ii) the worer s remtances provide an addional channel to induce financial development. This finding implies that policy maers should carry out economic policies to faster economic growth which, on the hand, leads to attract new remtances and, on the other hand, induces financial development. In addion, policy prescriptions in these countries should include new instruments to attract more remtances to the home countries and, in turn, this deepens financial sector. References Acosta, P.A., Baerg, N.R. and Mandelman, F.S., 2009, Financial development, remtances, and real exchange rate appreciation, Federal Reserve Ban of Atlanta Economics Review, 94 (), -2. Adams, R. and Page, J., 2005, Do international migration and remtances reduce poverty in developing countries? World Development, 33 (0), Aggarwal, R., Demirguc-Kunt, A. and Peria, M., 2006, Do worers' remtances promote financial development? World Ban Policy Research Woring Paper 3957, World Ban, Washington, D.C. Amuedo-Dorantes, C. and Pozo, S., 2004, Worers' remtances and the real exchange rate: A paradox of gifts, World Development, 32 (8),

7 The Empirical Economics Letters, 0(5): (May 20) 455 Brown, S.S., 2006, Can remtances spur development? A crical survey, International Studies Review, 8, Chami, R. and et.al., 2008, Macroeconomic consequences of remtances, Occasional Paper No. 259, International Monetary Fund. El-Saa, M. and McNabb, R., 999, The macroeconomic determinants of migrant remtances, World Development, 27, Giuliano, P. and Ruiz Arranz, M., 2009, Remtances, financial development, and growth, Journal of Development Economics, 90 (), Gupta, S., Pattillo, C. and Wagh, S., 2009, Effect of remtances on poverty and financial development in sub-saharan Africa, World Development, 37 (), Hnatovsa, V. and Loayza, N., Volatily and growth. World Ban Woring Paper No Im, K.S., Pesaran, M.H. and Shin, Y., 2003,Testing for un roots in heterogeneous panels, Journal of Econometrics, 5, Lartey, E., Mandelman, F. and Acosta, P., 2008, Remtances, exchange rate regimes, and the Dutch disease: A panel data analysis, Federal Reserve Ban of Atlanta Woring Paper, Levin, A., Lin, C. and Chu, C.J., Un root tests in panel data: Asymptotic and finesample properties. Journal of Econometrics 08, 24. Lopez, H., Molina, L. and Bussolo, M., 2007, Remtances and real exchange rate, World Ban Policy Research Woring Paper 423, Washington: D.C. Patric, H.T., 966, Financial development and economic growth in underdeveloped countries, Economic Development and Cultural Change, 4, Pedroni, P., 999, Crical values for cointegration tests in heterogeneous panels wh multiple regressors, Oxford Bulletin of Economics and Statistics, 6, Pedroni, P., 2000, Fully modified OLS for heterogeneous cointegrated panels, Advances in Econometrics, 5, Pedroni, P., 200. Purchasing power pary tests in cointegrated panels, Review of Economics and Statistics, 83, Rao, B.B. and Hassan, G.M., 200. A panel data analysis of the growth effects of remtances, Economic Modelling, doi:0.06/j, econmod, (forthcoming). Sayan, S., 2004, Guest worers remtances and output fluctuations in host and home countries, Emerging Marets Finance and Trade, 40, 68 8.

8 The Empirical Economics Letters, 0(5): (May 20) 456 Trew, A., 2006, Finance and growth: A crical survey, The Economic Record, 82 (259), Vargas-Silva, C. and Huang, P., 2006, Macroeconomic determinants of worers' remtances: Host versus home country's economic condions, Journal of International Trade and Economic Development, 5(), World Ban, 2006, Global Economic Prospects, 2006: Economic implications of remtances and migration, Washington, DC: World Ban.