empirical findings and its discussion thereof. Section V presents conclusion with policy implication.

Transcription

1 Causal Relatioship betwee Foreig Capital Iflows ad Ecoomic Growth: Empirical Evidece from Idia Naraya Sethi Assistat Professor i Ecoomics, Dept. of Humaities ad Social Scieces, Natioal Istitute of Techology (NIT), Rourkela, Odisha, Pi , Idia, ABSTRACT The preset study attempts to examie the casual relatioship betwee foreig capital iflows ad ecoomic growth i Idia. Usig the pair-wise Grager causality test (1969), this paper specifically examies casual relatioship betwee foreig capital iflows ad ecoomic growth i Idia. The importat observatios emerge from pair-wise Grager causality test which shows there is the log-ru equilibrium relatioship is restored betwee the followig pairs of variables viz., ecoomic growth ad Foreig Direct Ivestmet (FDI), ecoomic growth ad Foreig Portfolio Ivestmet (FPI). Keywords: Private Foreig Capital Iflows, Macroecoomic Variable, Vector Autoregressio (VAR), Idia 1. INTRODUCTION Iteratioal capital ivestmet ca play a useful role i developmet by addig to the savigs of low ad middle icome developig coutries i order to icrease their pace of ivestmet. However, foreig ivestmet ca also prove uproductive to developig ecoomies by exposig them to disruptios ad distortios from abroad, ad by subjectig them to surges of capital iflows or massive outflows of capital flight. Durig 1997 to 2001 the capital movemet to developig coutries has declied, but icreased margially i Iteratioal capital flow ca help developig ecoomies spread the beefit, whe the flows are steady ad do ot udermie the stability of fiacial system. Though capital accout liberalizatio i Idia bega oly i , Idia has history of capital iflows i the form of exteral assistace (durig 1950s), exteral assistace ad foreig ivestmet (durig 1960s), exteral assistace, foreig ivestmet, Exteral Commercial Borrowig (ECB), NRI deposits ad other capital sice the 1970s. Ecoomic growth is a log-term pheomeo: its casual factors are best studied over a period of time. Therefore the preset study restricts itself to the impact of capital iflows o ecoomic growth i the after liberalizatio period 1990s. The hypothesis i the study is based o the experiece of developig coutries o the growth related experiece with capital iflow. The ecoomic growth of ay coutry is fiaced either by its domestic savigs or the foreig savig that flow ito the coutry. Though foreig capital flows ito the coutry i the form of aid, Exteral Commercial Borrowig (ECB), ad NRI deposit, it does ot cotribute much towards Idia s capital formatio ad ecoomic growth. After 1993, whe the capital accout is partially liberalized, capital iflows have cotributed toward the idustrial productio ad ecoomic growth. The preset paper is divided five sectios icludig itroductio. Sectio II describes the review of earlier theoretical ad empirical literatures. Sectio III reports the data ad methodology. Sectio IV presets the empirical fidigs ad its discussio thereof. Sectio V presets coclusio with policy implicatio. 2. CAPITAL FLOWS AND GROWTH: LITERATURE SURVEY There have bee large umber empirical ad theoretical studies i the recet years o capital iflows ad their impact o macroecoomic variables. Recetly, Alfaro et al. (2005) examie the empirical role of differet explaatios for the lack of flows of capital from rich to poor coutries the Lucas Paradox". The theoretical explaatios iclude differeces i fudametals across coutries ad capital market imperfectios. They show that durig low istitutioal quality is the leadig explaatio. This study emphasizes the role of istitutios for achievig higher levels of icome, but remais silet o the specific mechaisms. The results idicate that foreig ivestmet might be a chael through which istitutios affect log-ru developmet. Aother importat study by Alfaro et. al. (2002) examie the various liks amog FDI, fiacial markets ad ecoomic growth. The empirical aalysis usig crosscoutry data betwee 1975 to 1995 shows that FDI aloe plays a ambiguous role i cotributig to ecoomic growth. However, coutries with well-developed fiacial market gai sigificatly from FDI. Studies o capital flows ito Idia ad its impact i macroecoomic variables have bee aalyzed by Kohli, (2003), Chakraborty (2001 & 2003), ad Dua ad Se (2006). Kohli (2003) examies capital flows o macroecoomic variables such as exchage rates, iterest rates of foreig exchage reserves, domestic moetary coditio ad fiacial system i Idia durig the period 1986 to She cocludes a iflow of foreig capital has a sigificat impact o domestic moey supply, stock market growth, liquidity, ad volatility. Correlatio betwee domestic ad foreig fiacial market highlights Idia s vulerability to exteral fiacial shocks. Chakraborty (2003) aalyses the fiacial crisis like East Asia crisis of ad the Mexica crisis of She uses the vector auto regressio (VAR) method to 65

2 examie the exteral shock geerated by capital iflows led to appreciatio i the real exchage rate as observed i the East Asia ad Lati America coutries. She fids from the impulse respose of the aalysis which reflects the fact that the impact of iflows of foreig capital o the real exchage rate durig the liberalized regime i Idia was differet from that observed i East Asia ad Lati America. Studies relatig to capital flows ad its impact o ecoomic growth i Idia have carried out by Ragaraja, (2001), Sethi ad Pataik (2007). Oe of the studies by Ragraja (2000) ivestigates the capital flows ito Idia ad its impact o the capital formatio ad ecoomic growth takig ito the variable as et private capital flows, et direct ivestmet, et official flows, et portfolio ivestmet ad other et ivestmets durig 1992 to He argues capital flows ca be promoted purely by exteral factors which may ted to be less sustaiable tha those iduced by domestic factors. Both capital iflows ad outflows, whe they are large ad sudde, have importat implicatio for ecoomies. He cocludes that the capital accout liberalizatio is ot a discrete evet. Sethi ad Pataik (2007) examie the impact of iteratioal capital flows o Idia s fiacial markets ad ecoomic growth. Usig mothly time series data from April 1995 to July 2005, they foud that Foreig Direct Ivestmet (FDI) positively affects the ecoomic growth, while Foreig Istitutioal Ivestmet (FII) egatively affects the ecoomic growth i Idia. test (1969) is used. However, the o-statioary ature of most series data ad the eed for avoidig the problem of spurious osese regressio calls for the examiatio of their statioary property. I first stage, Statioary of series o each variable is examied usig both Dickey-Fuller test ad Augmeted Dickey-Fuller tests. The Dickey-Fuller test ivolves estimatig regressio equatio ad carryig out the hypothesis test. To show the Dickey-Fuller (DF) test, the AR (1) process is show. Y t = α+ ρ.y t-1 +ε t.. (1.1) Where α ad ρ are parameters ad ε t is a white oise. Y is statioary, if -1<ρ<1. if ρ= 1, y is o statioary. if the absolute value of ρ is greater tha oe (ρ > 1), the series is explosive. Therefore, the hypothesis of a statioary series ivolves i whether the absolute value of ρ is strictly less tha o (ρ < 1). The test is carried out by estimatig a equatio with Y t-1 subtracted from both sides of equatios. Y t = α + γ Y t-1 + ε t (1.2) Where, γ = ρ 1 ad the ull ad alterative hypothesis are H 0: γ = 0 H 1: γ >1 I the coclusio of the above literature, we fid that, capital flows has sigificat impact o some macroecoomic variables i Idia. Also the capital flows betwee the coutries reduce the cost of capital, icrease ivestmet ad raise output. At a deeper level, however, it suggests that the experiece of growth ehacig effects of capital iflows has bee varied across coutries. 3. DATA AND METHODOLOGY 3.1 Descriptios Variables ad Sources of the Data The data for the study have bee collected from the secodary source such as Hadbook of Statistics i Idia Ecoomy (RBI) ad Iteratioal Fiacial Statistics (IFS), (IMF). The mothly data have bee take for the period from 1995:04 to 2011:07. The data of the study are private foreig capital iflows (FINV), Foreig Direct Ivestmet (FDI), Foreig Portfolio Ivestmet (FPI), Foreig Istitutioal Ivestmet (FII) ad idex of idustrial productio (IIP). The Idex of Idustrial Productio (IIP) has take as the proxy of GDP, though my study is based o mothly time series data, the mothly data of GDP is ot available. The period of study is costraied due to the uavailability of data after the liberalizatio period from So, the period of the study has bee take from 1995:04 to 2011: Methodology To examie the impact of capital flows o ecoomic growth i Idia, pair-wise Grager causality The t-statistics uder the ull hypothesis of a uit root does ot have the covetioal t-distributio. Dickey- Fuller (1979) shows that the distributio is o-stadard, ad simulated critical values for the selected sample. Later Mackio (1991) geeralizes the critical values for ay sample size by implemetig a much larger set of simulatios. Oe advatage of ADF is that it corrects for higher order serial correlatio by addig lagged differece term o the right had side. If the simple uit root test is valid oly if the series is a AR(1) process. Oe of the importat assumptios of DF test is that error terms are ucorrelated, homoscedastic as well as idetically ad idepedetly distributed (iid). Y t = α +γy t-1 +δ 1 Y t-1 + δ 2 Y t δ p Y t-p + ε t. (1.3) This augmeted specificatio is the tested for H 0: γ = 0 H 1: γ >1 I fial stage Grager s test causality is also kow as Weier-Grager test, sice its origi has bee traced to Wieer. To explai Grager test, we will cosider two statioary processes amely Y t ad X t are cosidered. The test ivolves estimatig two regressios amely, 66

3 Y t = X t = i X t-i + i 1 j 1 i X t-i + i 1 j 1 j Y t-j + u 1t...(1.4) j +Y t-j +u 2t...(1.5) Equatio (1.4) ad (1.5) respectively postulate that curret Y is related to past values of itself as well as that of X ad a similar relatio for X. I the above equatios, α s, β s ad δ s are parameters. I this cotext it is possible to distiguish three cases: Uidirectioal causality from X to Y is idicated if the estimated coefficiets o the lagged X i equatio (1.4) are statistically differet from zero as a group (i.e. α i 0) ad set of estimated coefficiets o the lagged Y i (equatio (1.5)) is ot statistically differet from 0 (i.e. δ j =0). Feedback or bi-directioal causality is suggested whe X causes Y ad Y causes X. I this case, the set of coefficiet of X ad Y are statistically sigificat differet from zero i both regressios. Fially, idepedece is suggested whe X does ot cause Y ad Y does ot cause X that is, the set of coefficiets are ot statistically sigificat i both the regressios. 4. EMPIRICAL ANALYSIS This sectio of the study discusses the impact of private foreig capital iflows o ecoomic growth usig pair-wise co itegratio test ad pair-wise Grager Causality test. The descriptive statistics of the variables are metioed i table 3. Uit root test i table 4 shows that oe of the variables are statioary at level, but they are becomig statioary o first differece. 4.1 Grager Causality Test Sice the reliability of results of the Grager causality test depeds o whether the variables are statioary or ot, we first tested uit root of the variables usig ADF ad DF test. The result of the uit root test is reported i table 2. It shows that all the variables are statioary o first differece. It is well-kow that Grager causality test is sesitive to the choice of lag legth. To avoid this problem, as oted i Eders (1995), we have applied Akaike iformatio criterio to choose the optimum lag legth. The results of pair-wise Grager Causality tests are reported i table 3. Major observatios are discussed here. The most importat observatio is that Ecoomic growth (IIP) Grager causes FDI ad FPI. This has relevace for the ecoomic policy after liberalizatio i Idia. It implies is that the past iformatio o ecoomic growth (IIP) improves the predictability of FDI. The above fidig, however, challeges this objective. We further observe that there is a bi-directioal casual relatioship amog these variables RFII ad RFDI, RFDI ad RFII, RFPI ad RFDI, RFDI ad RFPI, RIIP ad RFDI, ad RFDI, RIIP. But iterestigly, RIIP has a uidirectioal casual relatioship with RFPI ad RFII. It explais that the soud ecoomic growth of the coutry attracts additioal capital iflows (FPI, FII), whereas FII ad FPI have o casual relatioship with IIP. It meas that the high iflows of capital have o positive impact o ecoomic growth. This result suggests that, i the post reform period, istability i the tred behaviour of idex of idustrial productio (IIP) ca be explaied partly by the istability i the tred behaviour of the iflows of private foreig capital with some lagged effect. Table 1: Summary Statistics for FDI, FII, FPI ad IIP Descriptio FDI FII FPI IIP Sample Size Mea Media Maximum Miimum Std. Dev Skewess Kurtosis Jarque-Bera Probability Sum Sum Sq. Dev

4 Table 2: Uit Root Tests Results for FDI, FPI, FII ad IIP LEVELS Variable Without Tred With Tred DF ADF DF ADF FDI * (4) * (4) FII * (12) * (12) FPI * (12) * (12) IIP (4) * (6) FIRST DIFFERENCE Variable Without Tred With Tred DF ADF DF ADF RFDI * (4)* * (4)* RFII * (4)* * (4)* RFPI * (4)* * (4)* RIIP * (2)* * (4)* Notes: The critical values for uit root tests are -3.47, ad without tred ad -4.02, ad with tred. Figures i brackets agaist ADF statistics are the umbers of lags used to obtai white oise residuals ad these lags are selected usig AIC. *, **, *** imply sigificace at 1%, 5% ad 10% level respectively. Table 3: Pair wise Grager Causality Test for FDI, FPI, FII & IIP Depedet Explaatory Variable Variables m F-Statistics P-Value Remarks RFII RFII, RFDI causality from RFII RFDI RFDI RFDI, RFII causality from RFDI RFII RFPI RFPI, RFDI causality from RFPI RFDI RFDI RFDI, RFPI causality from RFDI RFPI RIIP RIIP, RFDI causality from RIIP RFDI RFDI RFDI, RIIP causality from RFDI RIIP RFPI RFPI, RFII No causality from RFPI RFII RFII RFII, RFPI No causality from RFII RFPI RIIP RIIP, RFII causality from RIIP RFII RFII RFII, RIIP No causality from RFII RIIP RIIP RIIP, RFPI causality from RIIP RFPI RFPI RFPI, RIIP No causality from RFPI RIIP Notes: (i) Optimum lag legths (m) are determied by miimizig the Akaike Iformatio criteria (AIC) by E-views package (ii) R: defies the percetage chage of the variables 5. CONCLUSION The preset study examied the impact of private foreig capital iflows o ecoomic growth usig pair wise Grager causality test. The causality test suggests a short ad log ru equilibrium relatioship betwee the variables like ecoomic growth ad foreig direct ivestmet ad ecoomic growth ad foreig portfolio ivestmet ad vice-versa. However, our empirical fidigs strogly show that there is dyamic short ad log equilibrium relatioship betwee variables durig the study period from 1995:04 to 2011:07. The study also fids that private foreig capital iflows have a positive ad direct impact o ecoomic growth. I other words, for the soud ecoomic growth of a coutry attracts additioal private foreig capital iflows. REFERENCES [1] Alfaro Laura., Kalemli-Ozca Sebem.,ad Volosovych Vadym.,(2005) Why does't Capital Flow from Rich to Poor Coutries? A Empirical Ivestigatio, Harvard Busiess School. [2] Chakrabarti, Rajesh., (2001) FII Flows to Idia: Nature ad Causes, Moey ad Fiace, 2 (7). 68

5 [3] Chakraborty, Idrai., (2003) Liberalizatio of Capital Flows ad the Real Exchage Rate i Idia: A VAR Aalysis, CDS Workig Paper, No [4] Chakraborty, Idrai., (2001) Ecoomic Reforms, Capital Iflows ad Macro Ecoomic Impact i Idia, CDS Workig Paper, No.311. [5] Chitre, Vikas., (1996) Foreig Capital Flows ad Fiacial Market i Idia, Joural of Foreig Exchage ad Iteratioal Fiace, 4, [6] Dicky, D.A. ad W.A Fuller., (1981) Like hood Ratio Statistics for Autoregressive Time Series with a Uit Root, Ecoometrica, 49, [7] Dua, Pami ad Se, Partha., (2006) Capital Flow Volatility ad Exchage Rates: The Case of Idia, Cetre for Developmet Ecoomics, Workig Paper No. 144, Departmet of Ecoomics, Delhi School of Ecoomics, Delhi Uiversity, New Delhi. [8] Egle, R.F. ad Grager, C.W.J., (1987) Coitegratio ad Error Correctio: Represetatio, Estimatio ad Testig", Ecoometrica, 55. [9] Khaa, Sushil., (2002) Has Idia Gaied from Capital Accout Liberalizatio? Private Capital Flows ad Idia Ecoomy i the 1990 s, Paper Preseted at the IDEAS Coferece, Iteratioal Moey ad Developig Coutries, Dec [10] Kohli, Reu., (2003) Capital Flows ad Domestic Fiacial Sector i Idia, Ecoomic Political Weekly, 22, [11] Mazumdar, T., (2005) Capital Flows ito Idia: Implicatios for its Ecoomic Growth Ecoomic Political Weekly, May. [12] Ragaraja, C., (2000) Capital Flows: Aother Look, Ecoomic Political Weekly, 9, [13] Reserve Bak of Idia, RBI (2011) Report o Currecy ad Fiace, Govt. of Idia, Idia [14] Sethi, Naraya ad Pataik, Uma Shakar K. (2007) Iteratioal Capital Flows o Idia s Ecoomic Growth-I View of Chagig Fiacial Market, The Idia Joural of Ecoomics, 45 (348). 69