COINTEGRATION-ERROR CORRECTION MODELING OF CAPITAL FLIGHT AND ENDOGENOUS GROWTH IN NIGERIA

Department of Economics & a Ph.D Scholar, Ebonyi State University, Abakaliki, Nigeria
Journal E-mails: martinslibrary1@gmail.com

Abstract
This study employs Error Correction Mechanism (ECM) to investigate the speed of adjustment between capital flight and economic growth in Nigeria using annual time series data spanning from 1970 – 2010.
The analysis begins with testing the stationary status of the time series by employing Augmented Dickey Fuller to confirm the number of integrating order. The study harvested that for most of the periods, capital flight estimates had positive sign, indicating that residents consistently took capital out of Nigeria. The study further documented that exchange rate volatility and high inflation rate are the  important means through which capital flight is effected in Nigeria, with evidences that confirmed the existence of financial revolving door relationship between capital flight and external indebtedness in Nigeria. The study, therefore, submits inter-alia repatriating of flight capital to boost the growth initiatives with selective controls on capital outflow, changes in Nigeria tax laws, and a bias toward poor wages.

Keywords: Capital Flight, Economic Growth, Cointegration and Nigeria.

Section I
Introduction
One of the crucial economic maladies that confronted most developing countries in the early 1970’s is the way capital flew from less developed nations to advanced nations of the world. In Nigeria for instance, London-based Tax Justice Network (TJN), recently, released a report totaling Nigeria’s Real capital flight (from 1970 to 2010) at a whopping $233.9 billion (Henry, 2012). However, they spoke against the backdrop of a report by the London-based Tax Justice Network (TJN) that Nigeria’s capital flight and investment income stood at $306.2 billion in 2011, about 40 times the nation’s total external debt of $7.9 billion. Since 1970s, it appears that private elites in developing world countries especially Nigeria, have been able to accumulate at least $7.3tn-9.3tn of offshore wealth, even while many of their public sectors were borrowing themselves into bankruptcy, enduring agonizing structural adjustment and low growth, and holding fire sales for public assets. Yet, Nigeria loses more than N2 trillion annually to capital flight following inability of indigenous ship owners to fully participate in crude oil exports (Okonjo-Iweala, 2012).

In a specific term, capital flight refers to any illicit movement of capital away from a domestic to a foreign economy. It is normally done in such manners that circumvent the regulatory purview of the domestic authorities. In the contemporary literature of development economies, there has been increasing attention to the notion of capital flight. Many analysts have attributed sluggish economic growth and persistent balance of payments deficits in most developing counties to capital flight (Ajayi, 1996). Capital flight is particularly a serious and an important concern for a developing country like Nigeria. First, capital is scarce in the developing world, so capital flight contributes to worsening the capital scarcity problem. In addition, it also restricts the capacity and ability of affected countries to mobilize domestic resources and access foreign capital necessary to finance economic growth and development.
Consequently, capital flight can contribute to the retardation of economic growth and development of developing countries. Second, capital flight can lead to a negative feedback. Because of the resulting tightening of capital constraints and the possibility of being cut off from foreign capital, even more capital flight could occur and consequently, progressive economic policies become more difficult to implement and raising social conditions a heavier burden to solve.

Thus, an estimate of capital flight at this period is important in order to know the relationship existing between capital flight and growth in Nigeria.  While many studies have been done on the topic, very few of these studies have been undertaken in relation to investment by Nigerians themselves. The studies of Ajayi (1996) which covered the period between 1970 and 1989 need a revisit. The studies of Onwudoukit (2001) did not provide any estimate and that of Lawanson (2007), was basically on capital flight with no relation to any other economic variable. These studies provided estimates to show the impact of trade misinvoicing or trade faking.   The producer price compounded figures for trade misinvoicing were put at $316,888 million ($316.9 billion) and $436,092.3 million ($436 billion) by Morgan Trust and World Bank respectively as at 2009 using the residual methods (World Bank, 2010). These figures are at variance with Collier’s, and need some clarifications. However, none of these studies empirically studied capital flight with its impacts on investment in the domestic economy.

Therefore, the resurgence of capital flight in recent times is related to a paradoxical situation of high accumulation of external debt by developing countries on the one hand and the acquisition of foreign assets by the citizens of the heavily indebted countries on the other. Consequently, interest is shown on capital flight at the policy to bring about reversal of capital flight as an opportunity not only to improve on the external liability situation of the economy but also to promote growth in Nigeria.

The remaining part of this study is divided into four sections. Section II deals with theoretical and empirical review. Section III highlights the methodological issues, section IV presents and analyses the empirical evidence while section V concludes the entire study.

Section II
Theoretical and Empirical Review
2.1 Theoretical Review
Capital flight theory, traditional, received only a scanty attention until recently. However, one basic theory upon which this study is leaned on is the portfolio theory of capital flight. Discussions on portfolio theory elicits high level of esotericism, when they tend to become dynamic, as static models and one country assumption can be easily understood and assimilated. The considerations of currency, real effective exchange rates, possibility of foreign or home bias, influences by the level of risks and real rates of return makes matters to be slightly complicated (Tille and Wincoop, 2007). However, the portfolio approach to international flows and flights of capital have been accepted as the most popular (Obstefeld, 2004). In the analyses of many of the flights of capital investigated, the portfolio approaches seem to have gained upper hand. The choice depends so much on the choice of investors who choose where to hold their wealth, either at home or overseas. The choice of either of these is influenced a lot by the risk and return trade off and other considerations.

Kraay and Ventura (2003) in their analyses, grouped the drivers of international portfolio flows into two namely: portfolio growth and portfolio reallocation models. The portfolio growth components are defined as increases in the national savings that lead to capital outflows which equals to the rise in national savings times portfolio share of foreign assets. The second one is active portfolio reallocation of wealth across assets. Capital outflows that relate to portfolio reallocation reflect a change in portfolio shares away from passive portfolio, since changes in assets price affect portfolio shares without any asset trade- a dimension known as passive portfolio management. The above scenario abandons the other relative factors that are involved in the inflow and outflow on the exchange rates and equity prices. An appreciable increase in the inflow of resources will undoubtedly affect the real effective exchange rate and the prices of the available equity stocks at home. This is the current position with our stock market and the relationship with the rate of exchange in its bubble days at about 2007 and 2008. This implies that the inflows of foreign portfolio funds into the capital market need to be monitored in order for it not to result in capital flight in the nearest future.

One significant aspect of this theory has been argued by several studies. This argument has been seen in more than one way. So important studies have been those of Lane and Millesi-Ferretti (2004) and Obstfeld (2004) who have called for the continuous use of portfolio approach in the explanation of countries of open economy dynamics. Though the studies of Kouri (1976) and Dooley and Isard (1982) did not support the use of portfolio approach initially with macroeconomic foundations as a result of lack of empiricism, the campaign nevertheless received caution earlier in Obstfeld and Rogoff (2002) as to where to draw the line in dynamic first order open economy before the current position. In Deveruox and Saito (2006), it was found that the existence of nominal bonds and the portfolio composition of net foreign assets is an essential element and a significant cause of capital flows between countries.  When investors adjust their gross positions in each currency’s bonds, countries can achieve an optimally hedged change in their net foreign assets (or their capital account), thus facilitating international capital flows.
2.2 Empirical Review
Not a few studies have been conducted to investigate the link between capital flight and economic performance in developing countries. However, the review undertaken here will be selective rather than been exhaustive. For instance, Lensink, Hermes and Murinde (2009) in their cross-sectional examination of the link between political risk and capital flight for a number of developing countries concluded that no matter how capital flight is defined or measured, political risk factors has a significant role to play in the determination of capital flight where no other macroeconomic variables are considered. Fatehi (1994) analyzed the impact of political disturbances on capital flight in 17 Latin American countries. He utilized a stepwise multiple regression analysis on data between 1950 and 1982. He concluded that political disturbances in some of those countries have effects on capital flight from these countries.

Forgha (2008) and Valeria Gusarova (2009) studing Cameroon and some developing nations respectively observed that capital flight adversely impact real economic growth.  Beja (2006) notes that with capital flight presents the possibility of cutting off a nation from external sources of funds. Consequently, it becomes more difficult to implement economic policies, and improving the social conditions of people also becomes more difficult.  Ajilore (2010) and De Boyrie (2011) observed that trade faking and mis-invoicing account majorly for capital flight in selected African countries including Nigeria and hinder long-term economic growth. Ayadi (2008) found interest differential and exchange rate depreciation significant causes of capital flight in Nigeria and concluded that capital flight is depriving Nigerian economy of substantial and critical financial resources needed for investment and building of social capital among others. Kosarev (2000) identified capital export as a normal economic phenomenon which does not affect the economy significantly from global perspective, while capital flight presents a danger and leads to the impoverishment of the economy.

Hermes and Lensink, (2000) and Lensink et al., (2000) in their empirical studies discovered that political instability in Africa is associated with greater capital flight whilst democracy and political freedom tend to reduce the incidence of capital flight. These together with weaknesses in the institutions for protecting property rights and incessant political unrest and associated general sense of insecurity to life and property tend to encourage capital flight in Nigeria. In his empirical study, Boyce (1992) identified 2 sets of bi-directional causality between external debt and capital flight, leading to the categorization of causal linkages into: debt driven capital flight; debt-fuelled capital flight; flight-driven external borrowing; and flight fuelled external borrowing. Contrary to the findings of Boyce, Ajayi (1996) found no evidence of causal links (in any direction) between external debt and capital flight. However, Collier et al. (2001), in a cross-sectional study, which includes some African countries, found evidence of debt-fuelled capital flight. Such finding is not surprising, as it must have been largely influenced by the presence of non-African countries in the sample where evidence of debt fuelled capital flight and flight-fuelled external borrowing had been reported earlier. Another major determinant of capital flight is risk-adjusted returns to investment. Certain studies have demonstrated a linkage between risk-adjusted returns to investment and capital
flight.This is argued on the assumption that investors attempt to maximize profits by diversifying their portfolios between foreign and domestic investments based on the relative risk-adjusted rate of return abroad and at home. Ndikumana and Boyce (2002 in their empirical studies, used exchange rate volatility, interest rate differential between home and abroad, and a host of survey-based measures of institutional investor risk perceptions to explain the concept of risk-adjusted returns to investment. Ndikumana and Boyce (2002) including Hermes and Lensink, (2009); Murinde et al., (1996); Nyoni, (2008); Ng’eno, (2000), which used interest rates as an explanatory variable in their models found no statistically significant relationship between interest rates and capital flight. However Murinde, et al., (1996); Hermes and Lensink, (2000) and Lensink et al., (2009), which used exchange rate indicators as an indicator of risk-adjusted returns found some evidence of the link between exchange rate overvaluation and capital flight in Nigeria.

Section III
Methodological Issue
3.1 Theoretical model
In order to capture the precise link between capital flight and endogenous growth in Nigeria, this study adopts Vector Autoregressive (VAR) model. The term autoregressive is due to the appearance of the lagged value of the dependent variable on the right-hand side and the term vector is due to the fact that we are dealing with a vector of two or more variables. The model is based on two lags of each endogenous variable. In a VAR model, each variable is in turn explained by its own lagged value, plus current and present value of the remaining variables. The VAR model present all variables as dependent variables which have the dynamic power to reflect impact of random disturbance on the variables, thereby modeling every endogenous variable in the system as a function of the lagged value of all the endogenous variable in the system. The VAR model presented here is composed of seven variables, namely: Economic growth (GDP), Capital flight (CAPFL), Exchange rate (EXCHR), Inflation rate (INFR) External Debt Relief (EXDR), Foreign Direct Investment (FDI) and Openness (OPEN). Thus, this study adopts a VAR model of Abdul Majid (2007) as modified by the researcher as follows:
at = SAiat-1 + et                                                                                                                       (3.1)
Where:
at = is a column vector of observation at time t on all the variance in the model,
S = summation of exogenous variable at time t
Ai = x1-x7
at-1= lag of endogenous variable.
et = stochastic error term or innovation of shocks.

Following the modeling approach of Abdul Majid (2007), we can specify the model in its implicit functional form as:
GDPt = (CAPFL, EXCHR, INFR, EXDR, FDI, OPEN)                                                       (3.2)
Now, stating the model in an explicit stochastic form gives;
GDPt = a0 + a1GDPt-1 + a2CAPFL + a3EXCHR + a4INFR + a5EXDR + a6FDI + a7OPEN + Ut
                                                                                                                                                (3.3)                                                  
Taking logarithms of both sides of the equation; we have,
LogGDPt = a0 + a1logGDPt-1 + a2logCAPFL + a3logEXCHR + a4logINFR + a5logEXDR + a6logFDI + a7logOPEN + Ut                                                                                                        (3.4)
where ;
a0 = constant and a1 to a7 = coefficients,
logGDPt = log of real Gross Domestic Product (GDP) growth in time t,
logGDPt-1 = log of lagged GDP in one year,
logCAPFL = log of Capital Flight,
logEXCHR = log of  Exchange Rate,
logINFR = log of Inflation Rate’
logEXDR = log of External Debt Relief,
logFDI = log of Foreign Direct Investment,
logOPEN = log of Openness of the economy (measures the difference between export and import), and
Ut = Disturbance element.
Apriori, it is expected that a1>0, a2>0, a3>0, a4<0, a5<0, a6>0, a7>0.


3.2 Sources of Data and Estimation Procedure
Data used for this study are mainly secondary in nature and were sourced from the Statistical Bulletin of the Central Bank of Nigeria (CBN) and Annual Abstract of Statistic of the Bureau of Statistics (NBS). The Ordinary Least Square (OLS) technique is used to investigate the link between capital flight and economic performance in Nigeria. Regression model was adopted to know the effect of capital flight on growth in Nigeria within the period under study. Also, coefficient of determination (R2), T-statistic, F-statistic, and the Durbin Watson test were employed to evaluate the significance of the estimated parameters of the regression model. The study also attempted to test for the time series characteristics using Augmented Dickey Fuller (ADF) Unit Root Test, Co integration and Error Correction Modeling, all at 5 or 10 percent level of significance.










Section IV
Presentation and Analyses of Results
4.1 Presentation of Results
In presenting the empirical evidence, table 1 below shows the output for absolute regression.
Table 1: Short-run Log Linear Result of Impact of Capital Flight and Growth
Dependent Variable: LOG (GDP)
Method: Least Squares
Date: 10/28/12          Time: 19:09
Sample (adjusted): 1970 2010
Included observations: 40 after adjusting endpoints
Convergence achievement after 6 iterations                                                                                                                         
            Variable         Coefficient    Std. Error       t-Statistic       Prob.                          
    C                     9.874991    0.203714       49.47489       0.0000
              LOG(CAPFL)   163.9319  534.7683       0.306547       0.6782
              LOG(EXCHR  -16.14560  7.842118          -2.058832   0.0584
              LOG(INFR)      -0.026677 0.021395       -1.246858      0.6529
              LOG(EXD)                0.858685     0.096694       8.880416       0.0000
              LOG(FDI)      0.073393     0.031426       3.335451       0.0262
              LOG(OPEN)    82.84605   37.28148           2.222177    0.0414                       
R-squared                     0.945867    Mean Dependent var.                      113.3829
Adjusted R-squared                0.836793    S.D Dependent var.              98.38592
S.E. of regression        0.182537    Akaite info criterion                        40.67834
Sum squared resid       0.281349    Schwarz criterion                 40.48392
Log likelihood                        -32.43872    F-Statistic                             89.72738
Durbin-Watson stat    1.873416    Prob (F-Statistic)                  0.000000      
Inverted MA Roots         .89                                                                                  
Source : Extracted from E-views package 6.0 Output
Recall that the specified model is
GDPt = a0 + a1CAPFL + a2EXCHR + a3INFR + a4EXDR + a5FDI + a6OPEN + Ut
Thus, using the absolute values of all the variables, the estimated parameters of the short run regression model is:
GDP = 9.874991 + 163.9319CAPFL - 16.14560EXCHR -        0.026677INFR + 0.858685EXDR + 0.073393FDI + 82.84605OPEN                                                                                         (3.5)

The estimated model shows that there exist positive relationship between real GDP and the explanatory variables – capital flight, external debt, foreign direct investment and openness of the economy. This empirical evidence is in conformity with the theoretical expectation except exchange rate and inflation rate which is expected to be growth retarding. The estimated result revealed that a unit change in capital flight (CAPFL), external debt (EXDR), foreign direct investment (FDI), and openness of the economy (OPEN) will enhance economic activities in Nigeria by values of 163.93, 0.86, 0.07 and 82.85 percent respectively. Likewise, a one percent change in exchange rate and inflation rate will retard growth by 16.15 and 0.02 percent respectively. However, the t-statistic is used to test for individual significance of the estimated parameters (a1, a2, a3, a4, a5, a6). The result reveals that not all the parameters estimated are significant (e.g. capital flight and inflation rate), because their respective t-calculated values of 0.31 and 1.25 are less than the t-tabulated value of 2.04. Therefore, the null hypothesis is accepted in that sense. This suggests that capital flight and inflation rate have not contributed to economic growth in Nigeria within the period under study. The f-statistic is used to test for a simultaneous significance of all the estimated parameters and the result showed that they are all simultaneously significant. This is so because the f-calculated (89.73) is greater than the f-tabulated (2.74). The Durbin-Watson test showed that there is little or no presence of serial correlation in the residual as its value (1.87) is approximately equal to 2. Overall, the Coefficient of Determination (R2) which measures how well the sample regression line fits the data is considered quite high, about (0.945867) or 95 percent. This implies that about 95 percent of the regression model was explained by the explanatory power. Only, an infinitesimal of 5 percent was unexplained.

Thus, the econometric analysis of the link between capital flight and growth in Nigeria within the years under study have shown that all the variables under investigation except exchange rate and inflation rate have positive relationship with economic activities in Nigeria, but the effect is insignificant. In conclusion however, the null hypothesis is accepted which implies that capital flight and its proxies have no significant impact on economic growth in Nigeria during the periods under review.    

4.2 Unit Root Test Analysis
In the unit root test, the variables for our analysis were subjected into a single type of unit root test (Augmented Dickey Fuller - ADF), to determine whether there is a presence of unit root or the series are stationary. Here, we investigated the time series characteristics of the variables (GDP, CAPFL, EXCHR, INFR, EXDR, FDI and OPEN) of the model in this study. For a brief, a variable is said to be stationary when it has no unit root which is denoted in literature as I(0). A non-stationary variable can have one or more unit roots and it is denoted by I(d), d is the number of unit roots that the variable possesses and, by implication, the number of unit roots that the variable must be differenced in order to make it stationary. Similarly, if a time series has to be differenced twice (i.e. take the first difference of the first differences) to make it stationary, we call such a time series integrated of order 2 -  I(2).

As depicted in table 2 below, all the variables are stationary at the first difference for each of the forms of estimation excepting capital flight which is stationary at second difference for all the three forms of the random walk model. This implies that all the variables of interest are integrated of order one i.e. I(1), excluding capital flight which is integrated of order two i.e. I(2). See the summary of the unit root test as depicted in table 2 below:

Table 2: Summary of Augmented Dickey-Fuller Unit Root Test
Variables      1% Critical Value        5% Critical Value      10% Critical Value      ADF T-Statistic       Order
D(GDP)                   -3.6117                -2.939                      -2.6080                     -7.626370       I(1)
D(CAPFR)        -3.6171                      -2.9422                  -2.6092                         -4.349505   I(2)
D(EXCHR)        -3.6117                     -2.9399                  -2.6080                         -5.122170              I(1)
D(INFR)                  -3.6117                -2.9399                  -2.6080                         -6.728054    I(1)
D(EXD)                   -3.6117                -2.9399                  -2.6080                         -5.236530   I(1)
D(FDI)             -3.6117                        -2.9399                  -2.6080                         -5.362451   I(1)
D(0PEN)          -3.6117                       -2.9399                  -2.6080                         -5.043710              I(1)
Source: Author’s Computation

4.3 Cointegration Test: Long-Run Analysis
So far, we have assumed that all the variables are of the same order of integration i.e. I(2), in order to carry out further tests. We then run an OLS regression of the variables at levels and test for cointegration by testing that the residual is I(1). This is the long run dynamic. The unit root test for the residual is carried out as follows: Recall again, our specified model as,
GDPt = a0 + a1CAPFL + a2EXCHR + a3INFR + a4EXDR + a5FDI + a6OPEN + Ut
The residual series is generated from the estimated model as shown below:
GDP = 9.874991 + 163.9319CAPFL - 16.14560EXCHR -        0.026677INFR + 0.858685EXDR + 0.073393FDI + 82.84605OPEN                            
Here, through transformation in residual, we have;
Ut = GDP – (9.874991 + 163.9319CAPFL - 16.14560EXCHR -          0.026677INFR+ 0.858685EXDR + 0.073393FDI + 82.84605OPEN)                                               (3.6)

Thus, the ADF is used to test whether the residual is stationary or non-stationary. Since the  estimated Ut are based on the estimated cointegrating parameters, a1, a2, a3, a4,a5, a6, the ADF critical significance values are not quite appropriate (Engle and Granger, 1987). Therefore, the ADF test in the present context is otherwise known as Augmented Engle-Granger (AEG) test. The result from the analysis revealed that the residual Ut is stationaryat 5 percent and 10 percent critical level since the tau value is more negative than the critical values; the null hypothesis of no cointegration is therefore rejected (see table 3 below). In conclusion, the residuals from the regression of the model of GDP on the exogenous variables as specified above are integrated of order zero i.e. I(0); that is, they are stationary. This implies that the regression is not spurious even though individually the incorporated variables in the model are non stationary at levels but all are stationary at first difference excluding capital flight that is stationary at second difference. Hence, the estimated model shows the static or long-run function of the relationship between capital flight and economic growth in Nigeria.          

Table 3: Johansen’s Cointegration Result for  Model of GDP                                
Date: 10/28/12          Time: 09:17
Sample (adjusted):   1970 2010
Included observations:  40 after adjusting endpoints
Trend assumption:   Linear deterministic trend
Series: GDP CAPFL EXCHR INFR EXD FDI OPEN
Lags interval (in first differences):1 to 1
Unrestricted Cointegration Rank Test                                                                 
Hypothesized                                                Trace             5 percent         1 Percent                   
No. of CE(s)         Eigenvalue Statistic    Critical value Critical value                 
None**                 0.968256                        138.7621       120.33                        137.28
At most 1              0.598161                        126.3850       114.31                        121.91
At most 2              0.856334                          73.9359          64.59                          93.53
At most 3              0.597324                          42.7372          33.84                          65.84
At most 4              0.315398                          29.7283          22.98                          46.27           
At most 5                  0.457239                      34.7398          46.47                          72.68
At Most 6              0.545762                          42.9532          89.33                        108.52
*(**) denotes rejection of the hypothesis at the 5%(1%) level
Trace test indicates 1 cointegrating equation(s) at both 5% and 1% levels              
Source: E-views Computer result

Table 4: Parsimonious Error Correction Model
Dependent Variable: D(LN(GDP)
Method: Least Squares
Date: 10/28/12          Time: 16:18
Sample (adjusted): 1970 2010
Included observation: 40 after adjusting endpoints                                                                                                                                                      
Variable                  Coefficient       Std. Error            t-Statistic              Prob     
    C                                0.038397    0.014075             2.728038             0.5612
 D(LN(GDP(-1)))            -0.065942           0.016049            -4.108824             0.0017
D(LN(GDP(-2)))                    0.883840      1.905095             0.463935             0.5672
DLN(GDP(-3)))                    -0.330889      0.307877            -1.074743             0.2684
D(LN(CAPFL(-1)))  7.422882       3.312050             2.241174             0.0567
D(LN(CAPFL(-3)           -0.563216            0.170897            -3.295652             0.0026
D(LN(CAPFL(-5)))         -2.200368           12.43412            -0.176962             0.6203
D(LN(EXCHR)))      0.072310       0.017959             4.026336             0.0020
D(LN(EXCHR(-1))) 0.058776       0.024360             2.412846             0.3413
D(LN(EXCHR(-2)))       -0.022334            0.211316            -0.105690             0.3468
D(LN(EXCHR(-3))) 3.423728       2.345612             1.459631             0.5622
D(LN(INFR)))                       0.283062       0.063248             4.475430             0.2683
D(LN(INFR(-1)))           -2.325676            1.138334            -2.043052             0.0672
D(LN(INFR(-2)))           -4.348634            2.386634            -1.822078             0.6689
D(LN(INFR(-3)))           -0.338943            0.384367            -0.868686             0.3621
D(LN(EXD(-1))))           -2.364286            1.394683            -1.695214             0.0626
D(LN(EXD(-2)))                   1.637266      0.421636             3.883127             0.3672
D(LN(EXD(-3)))                  0.083682       0.025677             3.259026             0.3961
D(LN(FDI(-1)))                    0.867623       0.373624             2.322182             0.0673
D(LN(FDI(-2)))                    -4.337463      1.856882            -2.335885             0.0000
D(LN(FDI(-3)))                    -0.720367      0.289243            -2.490525             0.7324
D(LN(OPEN(-1))           -0.328401            0.187384            -1.752556             0.0568
D(LN(OPEN(-3)))    5.326732       3.863742             1.378645             0.1682
D(LN(OPEN(-5)))    2.867287       0.934656             3.067745             0.0767
 ECM(-1)                  -1.568358       0.066348           - 0.066348             0.0000
R-squared                  0.996726       Mean dependent var.                                 0.038735
Adjusted R-Squared            0.898463       S.D. dependent var.                                     0.094636
S.E. of regression     0.367842       Akaike info criterion                                 17.67843
Sum squared resid    17.36384       Schwarz criterion                           17.12685
Log likelihood                      -43.73925      F-Statistic                                         9.567283
Durbin-Watson Stat 2.489265       Prob. (F-Statistic)                           0.000000             
Source: E-views Computer result          

4.4 Result of Error Correction Mechanism
The estimated result of Table 4 reports the initial over-parameterized error correction of capital flight and economic activities in Nigeria. All the variables were lagged equally in this model. The result of parsimonious model as reported above indicates model parsimony. Thus, this result clearly shows a well defined error correction term, and indicates a feedback of 157 percent of the previous year’s disequilibrium from the long run capital flight elasticity of economic activities in Nigeria. The implication of this result is that both capital flight and its proxies maintained equilibrium with GDP through time. The effects of these disequilibria error corrections is not only large, but also have negative signs as expected. The strong significance of the coefficient of ECMt-1 supports our earlier assertion that GDP indeed cointegrates with capital flight in Nigeria.
However, we previously showed that all variables under consideration are cointegrated at 5 percent and 10 percent critical level, i.e. there is a long-run relationship among them. In the short-run, there may be disequilibrium in which the model, i.e.
Ut = GDP – (9.874991 + 163.9319CAPFL - 16.14560EXCHR -          0.026677INFR+ 0.858685EXDR + 0.073393FDI + 82.84605OPEN     )                                                                                               (3.7)
 is the “equilibrium error”. Therefore, the error term is used to show the short-run behavior of real GDP to its long-run values. We can now specify the ECM equation for this study as:
ΔGDPt = a0 + a1ΔCAPFLt + a2ΔEXCHRt + a3ΔINFRt + a4ΔEXDRt + a5ΔFDIt + a6ΔOPENt + a7ECMt-1 + Ԑt
                                                                                                                                                (3.8)
 Where;
Δ denotes the difference operator; Ԑt is the random error term, and
ECMt-1 = (GDP -  a0 – a1CAPFLt-1 – a2EXCHRt-1 – a3INFRt-1 – a4EXDRt-1 – a5FDIt-1 – a6OPENt-1),  that is, the one-period lagged value of the error from the cointegrating regression. The ECMt-1 equation above states that ΔGDP depends on change in the explanatory variables and also on equilibrium error term that determines the short-run behavior of the model. The ECMt-1 equation is estimated through the use of E-view 6.0 and the result extracted from the E-view output as reported in table 4 above. Since, ECMt-1 is positive (i.e. GDP is above its equilibrium value), a7ECMt-1 will need to be negative which will cause ΔGDPt to be negative.
Therefore,  leading GDPt to fall in period t. Thus, the absolute value of a7 (1.000) decides how quickly the equilibrium is restored i.e. Ut-1 is the mechanism that adjust to the long-run equilibrium by a unit of any distortion that may occur in the short-run. The estimated ECMt-1 equation above shows that the short-run changes in all the exogenous variables have positive and significant impact on the short-run changes in the endogenous variable GDP. Therefore, the estimated parameters -  a1 to a6 are the short-run marginal effect on economic activities in Nigeria.

Section V
Conclusion and Recommendation
5.1 Conclusion
 Our major task in this paper is to investigate empirically the impact of capital flight on economic performance in Nigeria using recent econometric tool such as cointegration and Error Correction Mechanism (ECM). First, we begin with the analysis of time series with stochastic non-stationary components by analyzing the unit root properties of the relevant series. The results clearly show that the tests fail to reject the null hypothesis that these variables are stationary and they are, indeed, integrated of order two, which is the highest order i.e. I(2).

Given the stationary status of the series, the cointegration equation was estimated. The evidence, however, shows that capital flight, exchange rate, inflation rate, external debt, foreign direct investment and measures of openness cointegrate with the Gross Domestic Product (GDP) in Nigeria. It was also evidence that neither exchange rate (EXCHR) nor inflation rate (INFR) series seems to exert positively with the GDP series, rather increase in exchange rate with regards to naira and inflation galloping encourages capital flight, and consequently retard growth. The existence of one cointegrating linear combination was, therefore, established which corresponds to a long run GDP function with respect to exogenous variables under consideration. Based on this, an error correction model was developed which was shown to be well specified relative to its own information set and capable of parsimoniously representing the data set.
Adopting a cointegrating and an error correction modeling strategy, the relationship between Nigeria’s GDP and capital flight were analyzed through a series of reduction from over parameterized model interrelating all the components of the GDP models. Thus, the estimates presented in this study suggest that high exchange rate and inflationary rate encourage capital flight, and hence retard economic activities in Nigeria.
5.2 Policy Recommendations.
Emanating from the result, for capital flight and its various components such as exchange rate, inflation rate, external debt relief, foreign direct investment and openness of the economy to have significant impact on economic growth in Nigeria, the following policy options are recommended: First, the huge estimates of capital flight suggest a huge potential for capital flight reversals. Efforts must be made towards the design and implementation of appropriate policy measures that would encourage flight capital to return to the country. Better economic reforms that will encourage the inflow of foreign capital should be made. The reform should thus be based on the need to encourage growth, exchange rate reduction, single-digit inflation and reverse the negative distributional effects of capital flight. Specific policies might include repatriation of flight capital to boost the growth initiatives with selective controls on capital outflow, changes in Nigeria tax laws, and a bias toward poor wages. More generally, a new overall strategy that would encourage Nigerians abroad to come back home and invest in the country is required. Repatriate

References

Abdul- Majid , A. (2007). Consequences of Capital Flight for Latin American Debtor Countries. In Lessard and Williamson, eds., Capital Flight and Third Wwld Debt. Washington, D.C., Institute for International Economics, 1987.

Ajayi, S. Ibi. (1997). “An Analysis of External Debt and Capital Flight in the Severely indebted Low Income Countries in Sub-Saharan Africa”, Research Department, International Monetary Fund Working Paper 68 (Washington DC).

Ajilore, O.T. (2005). “External Debt and Capital Flight in Nigeria: Is There a Revolving Door?. South African Journal of Economic and Management Sciences. Vol. 8 No.2. pp 47-56

Ayadi, K.. (2008). Capital Flight: Estimates, Issues, and Explanation, Princeton Studies in International Finance, No. 58.

Boyce, J.K. (1992). “The Revolving Door? External Debt and Capital Flight: A Philippine Case Study”, WorldDevelopment, Vol.20, No.3. pp 1342 - 1357

De Boyrie, T. (2011). Macroeconomic Determinants of Capital Flight: An Econometric Investigation. In Lessard and Williamson, eds. ,Capital Flight and Third World Debt. Washington, D.C.: Institute for International Economics.

Deveruox, N. and Saito, L.. (2006). The ‘Problem’ of Capital Flight: A Cautionary Note. The
World Economy, 12(2), 237-252.

Dooley, S. and Issard, L. (1970). Capital Flight from Latin America. World Development, Vol. 18, No. 1, January 1790 (forthcoming).

Engle, N. and Granger C. W. F. (1987). “Development in the Study of Cointegrated Economic Variables”. Oxford Bulletin of Economics and Statistics, Vol. 48, No.3, pp. 213-28.
Fatehi, K. (2004). Capital Flight from South Africa, 1980-2000. African Development and Poverty Reduction, Forum Paper.

Forgan, S. (2008). Comparative Aspects of the Magnitude and Determinants of Capital Flight in Six Sub-Saharan African Countries. Savings and Development Quarterly Review, 20(1), 61-78.

Henry, J. (2012),” The Guardian of London news report” London

Hermes, N and R. Lensink. (2000). “Capital Flight and the Uncertainties of Government Policies”, Economics Letters, Vol.71 No.3. pp 87 - 96

Kouri, R. (2000). Capital Flight from Tanzania. In: Ajayi, I. and Khan, M.S. (eds). External Debt and Capital Flight in Sub-Saharan Africa. Washington, D.C.: The IMF Institute., 238-264.

Koutsoyiannis, A. (2003), “Theory of Econometrics” 2nd edition, Palgrave, Hound mills’ basingtoke, New York.
Kraay, H. and Ventura, A.. (2003). Financial Liberalisation and Investment. Routledge, London.

Lane,.K. and Millesi-Ferretti, A. (2004). Public Debts and Private Assets: Explaining Capital
Flight from Sub-Saharan African Countries. Political Economy Research Institute, UMASS, Working Paper Series No. 32.

Lawanson A. Olayinka. (2007). Econometric analysis of capital flight from nigeria: a portfolio choice approach. Final Report Submitted to African Economic Research Consortium, Nairobi, KENYA

Lensink, R., N. Hermes, and Victor Murinde. (2009). “The Effect of Financial Liberalisation on Capital Flight in African Economies”, World Development, Vol.26 No.7. pp 1467- 1477

Lensink, Robert, Niels Hermes, and Victor Murinde. (2000). “Capital Flight and Political Risk.” Journal of International Money and Finance, 19: 73-92.

Ng’eno, N.K. (2000). “Capital Flight in Kenya”, in S.I. Ajayi and M.S. Khan (eds).. External Debt and Capital Flight in Sub-Saharan Africa. Washington, DC: The IMF Institute.

Nyoni, T (2000). “Capital Flight from Tanzania”, in S.I. Ajayi and M.S. Khan (eds).. External Debt and Capital Flight in Sub-Saharan Africa. Washington, DC: The IMF Institute.

Obstefeld, M. and Rogof, D. (2002). “The Revolving Door? External Debt and Capital Flight: A Philippine Case Study”, WorldDevelopment, Vol.20, No.3. pp 1342 - 1357

Okonjo-Iweala, N. (2012), “Presidential Retreat on Maritime Safety Held at the Presidential Villa.” Abuja.

Onwioduokit, E.A. (2001). “Capital Flight from Nigeria: An Empirical Re-Examination”, Paper Presented at WIDER Development Conference on Debt Relief, held at Helsinki, Finland 17-18 August.

Tille, D. and Wincoop, C. (2007), “The Size and Sustainability of Nigerian Current Account Deficits” International Monetary Fund Working Paper 01/87 (Washington DC).

World Bank (2010). “Case Study: Mexico” in Capital Flight and Third World Debt, edited by D.R. Lessard and J.Williamson (Washington DC: Institute for International Economics
Share on Google Plus

Declaimer - MARTINS LIBRARY

The publications and/or documents on this website are provided for general information purposes only. Your use of any of these sample documents is subjected to your own decision NB: Join our Social Media Network on Google Plus | Facebook | Twitter | Linkedin

READ RECENT UPDATES HERE