Asian Economic and Financial Review AN ERROR CORRECTION REPRESENTATION OF MARKET LIQUIDITY ECONOMIC GROWTH NEXUS IN NIGERIA: A RECENT EXPERIENCE

Size: px

Start display at page:

Download "Asian Economic and Financial Review AN ERROR CORRECTION REPRESENTATION OF MARKET LIQUIDITY ECONOMIC GROWTH NEXUS IN NIGERIA: A RECENT EXPERIENCE"

Janice Fowler
5 years ago
Views:

Asia Ecoomic ad Fiacial Review ISSN(e): 2222-6737/ISSN(p): 2305-2147 URL: www.aessweb.

1 Asia Ecoomic ad Fiacial Review ISSN(e): /ISSN(p): URL: AN ERROR CORRECTION REPRESENTATION OF MARKET LIQUIDITY ECONOMIC GROWTH NEXUS IN NIGERIA: A RECENT EXPERIENCE Kaie, A.I.N Olotu, McRollis Efe 2 1 School of Agricultural Scieces, Natioal Ope Uiversity of Nigeria, KM4 Kadua-Zaria Express Way, Rigachiku, Kadua, Nigeria 2 Departmets of Ecoomics, Novea Uiversity, Ogume, Delta State, Nigeria ABSTRACT The study examies the relatioship betwee market liquidity ad ecoomic growth i Nigeria over the period of usig time series data. The startig poit of our aalysis was to check for the time series properties of the uderlyig data usig the Augmeted Dickey Fuller (ADF) ad the Egle-Grager procedure. The existece of ostatioarity ad co-itegratio were established which ecessitated a ECM for the aalysis. But first, the popular Grager-Causality Test was coducted to ascertai whether Market Liquidity Grager-causes Ecoomic Growth or vice-versa. The empirical result from the Grager causality test idicates that market liquidity does ot gragercause GDP; either does GDP grager-cause market liquidity. From the result, Market Liquidity captured by its proxies of Tur-over Ratio, Market Capitalizatio ad Private Sector Credit are statistically sigificat i explaiig GDP. Based o the fidigs, we suggest the moetary authorities should focus o policies that will develop fiacial ifrastructure that will esure macroecoomic stability uder which ecoomic agets ca make decisios that promote rapid fiacial market developmet ad subsequet capital market performace. The problem of low awareess about the operatios of the capital markets i the ecoomy ca be reduced by educatig the populace about the opportuities available i the market. Keywords: Market liquidity, Tur-over ratio, Tradig volumes, Ecoomic growth. Cotributio/Origiality The study is oe of the very few which have ivestigated the market liquidity-ecoomic growth exus i Nigeria i recet time ad cotributes the first logical aalysis of a Error Correctio represetatio of the pheomea makig it differet i methodology ad i scope. 1. INTRODUCTION Theoretical uderpiig stipulates that for growth to evolve i ay cotemporary emergig ecoomy, adequate ivestmet is a sie qua o. This ivestmet i tur requires log-term fudig/capital a sceario which is oly possible if the fiacial markets are highly developed. By providig heterogeeous products which eable ecoomic agets to cope with ucertaities through hedgig, poolig, sharig, ad pricig risks, these fiacial markets as observed by Kremli ad Vlagaskovic (2010) sigificatly affect ecoomic growth by trasformig savigs ito ivestmet. I Africa, the market is highly illiquid, with very low tradig volumes ad turover ratios. This has serious implicatios for ecoomic activities, as Levie ad Zervos (1998) observed that, high market liquidity reduces Correspodig author DOI: /joural.aefr/ / ISSN(e): /ISSN(p):

2 risk through risk sharig, which ecourages savig ad ivestmet. Therefore, low liquidity implies limited opportuities for trasformig illiquid assets ito liquid assets, which ca hitherto costrai ecoomic activity. Liquidity broadly defied, is the ease ad speed with which agets ca covert assets ito purchasig power at agreed prices (Levie ad Zervos, 1998). To udertake this coversio, there are associated risks techically omeclated liquidity risks which arise due to ucertaities i covertig assets ito a medium of exchage. These ucertaities i tur, result from iformatio asymmetries ad trasactio costs which ihibit liquidity ad itesify liquidity risk. A liquid capital market is thus a market where it is relatively cheap ad easy to trade fiacial istrumets ad eable firms have permaet access to the capital ivested by iitial shareholders with low level of ucertaity about the timig ad settlemet of the trasactio. Sice most high-retur projects require log-term commitmet of capital, sice savers fid it puzzlig to give up cotrol of their savigs for log period, if the fiacial system does ot augmet the liquidity of log term ivestmets, less ivestmet is likely to occur i high-retur projects. I other words, to esure higher ecoomic growth, greater market liquidity ad efficiecy i service delivery is idispesable. This is why the Nigeria Stock Exchage, accordig to Cetral Bak of Nigeria (2007) commeced operatio o its Trade Alert, Trade Guaratee Fud Scheme ad its e-busiess Platform/iteret portal i the year That is ot all. It has bee opied i some quarters that the effect of market liquidity is ambiguous (Blide, 1993). Thus, cotrary to the sytheses above, savigs rate may rise or fall as market liquidity rises. As exhumed by Jappelli ad Ragoa (1994) amog others, i a model with physical capital exteralities, savig rates could fall eough, so that growth actually decelerates with greater liquidity by reducig the icetives of shareholders to udertake the costly task of moitorig maagers (Sleifer ad Vishy, 1986). It is o the backgroud of the disparity i the determiatio of liquidity risk by differet forces as it affects ecoomic growth that ecessitated this research to ivestigate the ature of the relatioship betwee market liquidity ad ecoomic growth i Nigeria ad ascertai if market liquidity cotribute to ecoomic growth i Nigeria. 2. MATERIALS AND METHODS 2.1. The Model The study employs the popular Grager causality test suggested by Grager (Egle ad Grager, 1987) where a urestricted equatio with lags of a particular variable is estimated ad the the joit sigificace of each variable is tested. To do this, we specify two equatios with ecoomic growth ad market liquidity as depedet ad their respectively j lags as idepedet variables as follows: GDPt jmlq jgdp U...(1 a) t j j i it j 1 j 1 MLQ jmlq jgdp U...(1 b) t t j t j 2t j 1 j 1 Where: GDP = Ecoomic Growth proxied by Gross Domestic Product MLQ = Market Liquidity proxied by Tur-Over Ratio. The decisio rule is to accept the ull hypotheses if the F-calculated is less tha the F-critical ad reject if otherwise at 5% level of sigificace. That is to say, we will accept the ull hypothesis that market liquidity does ot grager-cause GDP if F-calculated is less tha the F-critical. Also, i the secod hypothesis, we accept the ull hypothesis that GDP does ot grager-cause market liquidity if the F-calculated is less tha the F-critical ad reject if otherwise at 5% level of sigificace ad vice-versa. 110

3 Next,drawig ispiratios from the literature ad guarded by the objective, a Model of Market Liquidity ad Ecoomic Growth i Nigeria is specified as follows: Where: GDP f ( MCT, TOV, PSC, FCI, Ut...(2) GDP = Real Growth Domestic Product MCT = Market Capitalizatio TOV = Turover Ratio PSC = Private Sector Credit FCI = Foreig Capital Iflow Assumig a liear relatioship betwee the variables ad usig the expected sigs, the mathematical equatio of the above fuctio trasforms to: GDP 0 1MCT 2TOV 3PSC 4 FCI t Estimatio Procedure It has bee observed that results emaatig from most macroecoomic variables are likely to be spurious if the time series properties are ot examied. Cosequetly, prior to estimatio, basic tests will be coducted as precautio. The first step of this aalysis is to ivestigate the existece of o-statioarity usig ADF test (Iyoha ad Ekaem, 2002) give by the followig of equatio: Y Y Y et t 1 2t t 1 1 t 1 i 1 m...(4) Where: Y t = the chage i the logarithm of the time series. Y t 1 = the lagged values of the depedet variables m = chose to elimiate the autocorrelatio I the above if = 0, there is evidece of uit root. Where ay variable is foud to be o-statioary, it will be differeced. Similarly, the Egle-Grager procedure would be applied to check for the presece or otherwise of logterm relatioship. If the residual is statioary, the a log ru relatioship is established amog the variables. Thus we shall itroduce the Error Correctio Model to accout for the speed of adjustmet thus: GDP MCT... FCI...(5) t 0 1 t 4 t t t i It i 0 i 0 u = Error Correctio Represetatio 1 t 1 Where; = Coefficiet measurig the degree of error corrected The Dataset/Aalytic Software The data for this study was obtaied from the CBN (2013) ad the Iteratioal Fiacial Statistics (IFS) (2010) amog others ad spas the period The Microsoft Excel will be used to eter the data ad trasported via Lotus wks123 to Pc-Give (versio 8.0) where they will be aalyzed usig the OLS. 111

4 3. PRESENTATION AND ANALYSIS OF RESULTS 3.1. Result from Uit Roots ad Co-Itegratio Tests Usig Augmeted Dickey Fuller (ADF) test for the equatio, we have table A below: Table-A. Uit Roots Test ADF-test Variable DGDP DMCT DTOV DPSC DFCI Level st Differece ** ** ** ** ** 2d Differece,,,,,,,,,, NB** idicate sigificace at the 5% level. Source: Authors computatio usig Pc-Give (versio 8.0) via OLS.. The data sourced from CBN (2013) ad the IFS (2010) were keyed ito Microsoft Excel ad trasported via Lotus wks123 to Pc-Give The test shows that all the variables are itegrated of order oe. Give this uit root property, we implemeted the Egle-Grager co-itegratio test where the liear combiatio of GDP (the depedet variable) ad all the explaatory variables are ru i their level forms without the itercept. The result is displayed i table B below: Table-B. Result from Co-Itegratio Test for the Residuals t-adf Lag Critical value Residual Residual Residual Source: Authors computatio copied from Pc-Give (versio 8.0). The data sourced from CBN (2013) ad the IFS (2010) were keyed ito Microsoft Excel ad trasported via Lotus wks123 to Pc-Give From the table, oe of the t-adf exceeds the critical value (-1.952) at the 5% level. The result shows the presece of co itegratio because the residual obtaied from the liear combiatio of the variables were statioary while the series geeratig the residual are ot statioary. Due to this techical error, a Error Correctio Model (ECM) which will help to correct most of the errors iheret i the model becomes idispesable Result from Grager-Causality Estimate The results of the Grager-causality test betwee Market Liquidity ad GDP is preseted i table C below: Table-C. Directio of Causality betwee GDP ad MLQ Null Hypothesis F-value F- tab Decisio MLQ does ot grager cause GDP Accept GDP does ot grager cause MLQ Accept Source: Authors iterpretatio of computed result as idicated i paragraph two of subsectio 2.1 From the result, the F-calculated i the first hypothesis is while the F-critical is Sice the F- calculated is less tha the F-critical at 5% level of sigificace, we accept the ull hypothesis that market liquidity does ot Grager-cause GDP. Similarly, i the secod hypothesis, the F-critical exceeds the F-calculated thereby ecessitatig the acceptace of the ull hypothesis that GDP does ot Grager-cause market liquidity. This fidig implies a zero causal relatioship betwee both variables. 112

5 3.3. Results from Modelig GDP by OLS Havig ivestigated the time series properties of the variables, we estimated the ECM by OLS. For simplicity, the result is preseted i table D below: Table-D. Estimated Model Variable Coefficiet t-value PartRý Costat DMCT DTOV DPSC DFCI ECM_ Rý=0.7104; F(5, 21)=24.15[0.0000]; DW=2.19; RSS=5.29 Source: Authors computatio copied from Pc-Give (versio 8.0). The data sourced from CBN (2013) ad the IFS (2010) were keyed ito Microsoft Excel ad trasported via Lotus wks123 to Pc-Give The table above ca also be represeted i a cocise, compact form as: GDP MCT TOV PSC FCI ECM _1... * (3.79) (2.84) (-9.82) (-1.35) (-2.49) From the result, the overall goodess of fit of the model is commedable. This is show i the high R 2 value of This suggests that 71% of the variatio i GDP is explaied by all the explaatory variables take together. The F test is a joit test to ascertai if the R 2 is statistically sigificat. From the rule, F-calculated (5, 21)=24.15 exceeds F critical of 2.77 at the 5% level. Thus, the F-statistics sigificatly affirmed the positio of the R 2. The Durbi Watso test is employed to test for the presece or otherwise of autocorrelatio. Usig the rule of the thumb, if DW-statistics falls withi the rage of absolute value of two, there is o autocorrelatio. However, if it exceeds or is less tha the two, it implies the presece of autocorrelatio. Give the DW value of 2.19, we coclude the presece of slight autocorrelatio at 5% level of sigificace. From the result above, the idex of Market Capitalizatio possesses a positive sig ad a robust coefficiet of Judgig from the t-value of 3.79, the variable is statistically sigificat i explaiig GDP i Nigeria. Thus, a uit icrease i market capitalizatio brigs about 6.38 uits icrease i ecoomic growth all thigs beig equal. Similarly, Private Sector Credit also possesses a positive sig ad a coefficiet of From the result, the t- observed is 9.82 implyig that the variable is statistically sigificat i explaiig variatios i GDP. Thus, other thigs beig equal, a uit icrease i Private Sector Credit to the ecoomy brigs about icreases i GDP to the tue of 1.2uits. Although, the estimated coefficiet for Tur-over Ratio is fragile, the variable proves statistically sigificat i explaiig GDP i Nigeria judgig from the t-value. Thus, other thigs beig equal, a uit icrease i the variable leads to 0.04uit icreases i GDP. The oly variable employed for the study which fails to coform to apriori expectatio is the Foreig Capital Iflow. The variable did ot oly fail i the expected positive relatioship, it is also ot statistically sigificat. The ECM compoet satisfied the stadard regressio assumptios ad is also sigificat. The result which shows the speed of adjustmet to log-ru equilibrium idicates that at every regular iterval, about 41 percet of the error ecoutered is corrected i the model. 4. SUMMARY, CONCLUSION AND RECOMMENDATION The study examies the relatioship betwee market liquidity ad ecoomic growth i Nigeria over the period of usig time series data. The statig poit of our aalysis was to check for the time series properties of the 113

6 uderlyig data. Thus, uit roots ad coitegratio tests were coducted usig Augmeted Dickey Fuller (ADF) ad the Egle-Grager procedure. The existece of o-statioarity ad co-itegratio were established which ecessitated a ECM for the aalysis. But first, the popular Grager-Causality Test was coducted to ascertai whether Market Liquidity Grager-causes Ecoomic Growth or vice-versa. The empirical result from the Grager causality test idicates that market liquidity does ot grager-cause GDP; either does GDP grager-cause market liquidity. This implies the complete absece of a causal relatioship betwee both pheomea. The stage was set for the mai regressio. From the result, Market Liquidity captured by its proxies of Tur-over Ratio, Market Capitalizatio ad Private Sector Credit are statistically sigificat i explaiig GDP. Based o the fidigs of the study, we suggest the moetary authorities should focus o policies that will develop fiacial ifrastructure that will esure macroecoomic stability uder which ecoomic agets ca make decisios that promote rapid fiacial market developmet ad subsequet capital market performace, sice the back boe of a atio s ecoomy ceters o its fiacial system. Also, there is eed to etrech a high level of trasparecy i all facets of the market as obtaiable i the developed markets. Fially, the regulatory framework of the Nigeria fiacial market should be reviewed to attract public cofidece for the market. REFERENCES Blide, A., The hidde costs of stock market liquidity. Joural of Fiace Ecoomic, 34(5): CBN, Statistical bulleti ad statemet of accouts. Directorate of Research ad Statistics, 24: Cetral Bak of Nigeria, Capital markets dyamics i Nigeria: Structure, trasactio costs ad efficiecy. Research ad Statistics Departmet., Abuja. Cetral Bak of Nigeria. Egle, K.F. ad W.F. Grager, Co-itegratio ad error correctiio represetatio ad testig. Ecoometrica, 55(3): Iteratioal Fiacial Statistics (IFS), Available from Iyoha, M.A. ad T.O. Ekaem, Itroductio to ecoometrics. Bei City: March Publishers. Jappelli, T. ad M. Ragoa, Savig growth ad liquidity costraits. Quarterly Joural of Ecoomics, 109(1): Kremli, J. ad A.T.K. Vlagaskovic, Volatility of stock markets, ivestors diversificatio ad iteralizatio of market ucertaities i emergig ecoomies: Evidece from Uzbekista. Russia Ecoomic Review, 79(4): Levie, R. ad S. Zervos, Capital cotrol liberalizatio ad stock market developmet. Joural of Ecoomic Developmet, 19(2): Sleifer, A. ad Vishy, Large share holders ad corporate goverace. Joural of Public Fiace, 94(3): Views ad opiios expressed i this article are the views ad opiios of the authors, Asia Ecoomic ad Fiacial Review shall ot be resposible or aswerable for ay loss, damage or liability etc. caused i relatio to/arisig out of the use of the cotet. 114

AY Term 2 Mock Examination

AY Term 2 Mock Examination AY 206-7 Term 2 Mock Examiatio Date / Start Time Course Group Istructor 24 March 207 / 2 PM to 3:00 PM QF302 Ivestmet ad Fiacial Data Aalysis G Christopher Tig INSTRUCTIONS TO STUDENTS. This mock examiatio