Statistical Analysis of Interrelationship between Money Supply Exchange Rates and Government Expenditure Evidence from Pakistan

50 %
50 %
Information about Statistical Analysis of Interrelationship between Money Supply Exchange...

Published on January 6, 2016

Author: AtifAhmed14

Source: slideshare.net

1. A Statistical Analysis of Interrelationship between Money Supply, Exchange Rates and Government Expenditure: Evidence from Pakistan Atif Ahmed 9549 atif_contact11@yahoo.com Syed Muhammad Owais 9542 sm_live90@hotmail.com Muhammad Asad Ali 9624 asad6281@gmail.com Hafiz Wasif Kamal 9323 hafizwasif88@gmail.com Muhammad Zakariya Qazi 9580 zakqazi@hotmail.co.uk Research Method (Wed) Submitted To Tehseen Jawaid Fall Semester (2014)

2. 2 Acknowledgements First of all thanks to All Mighty Allah and then "We would like to thank our teacher, Mr. Tehseen Jawaid for the valuable advices and support which he has given to us on writing of this report. Regards

3. 3 Chapter 1

4. 4 1. Introduction Several researches have been conducted to study the impact of different macro-economic variables and their influence on government expenditure. By using different statistical tools researchers have examined that how money supply and exchange rate influence the government expenditure. Few other studies also conducted work on the quarterly time series data to examine the long run equilibrium association between the macroeconomic variables. Friedman (1978), Blackely (1986) Ram (1988) suggested that the increase in government revenue plays a vital role in increasing the government expenditure. Akhtar and Abbas (2002) drawn the conclusion that exchange rates might differ due to decrease in government expenditure. Guffey (1985) concluded that money is not neutral, because it is the monetary policy that heavily effects inflation which will again adversely affect the economic growth. Studying over 100 countries it is suggested by Landau (1983) that growth rate and government expenditure are negatively related. Grier and Tullock (1989) use pooled regression on five-year averaged data in 113 countries to analyze the relationship between cross-country growth and various macroeconomic variables. They find that the mean growth of government share of GDP generally has a negative impact on economic growth. This finding implies that an increase in the government size as measured by a share of government expenditures to GDP hampers economic growth. Barro (1990) also discovers the negative 0 10 20 30 40 50 60 70 80 90 100 Year 1972 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 ExchangeRates YEAR Exchange Rate Value Exchange Rate Value

5. 5 relationship between the size of government and economic growth. Miller and Russek (1997) indicate that debt- finance increases in government expenditure retarded growth. It is evident from previous researches that money supply has an impact on government expenditure as well as economic growth. By taking this hypothesis this research made an attempt to study these variables and their relationship in Pakistan economy in time period of 1970 to 2012. Also it will investigate that what role is played by exchange rate in government spending and will try to find out its nature. 0 10 20 30 40 50 60 Year 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 Moneysupply YEAR Money and quasi money (M2) as % of GDP

6. 6 Chapter 2

7. 7 2. Literature Review 2.1.Theoretical Background The theoretical relationship between government expenditure and money supply and exchange rates is very important to understand. According to monetarist theory it is assumed that changes in money supply exert a dominant influence on changing patterns of government expenditure. The public’s demand for money is another important part of the relationship between money supply and government expenditure1 . This should be kept in mind that if Rs. 1 bought yesterday what Rs. 2 bought today then there will be a need for Rs. 2 to sustain the purchasing pattern. That’s why the increased government expenditure is a false increase in money supply because it is the deficit that government needs to finance, a reserve of money is there for loans for the government from State Bank of Pakistan and other financial institutions. Now, the exchange rate is also a major macroeconomic variable that if left ungoverned can be very threatening in fiscal policy making2 . Literature that was analyzed for this research paper shows little or no sign of exchange rates impacting government expenditure, which is why this research paper might throw some light on the nature of relationship between exchange rates and government expenditure. 2.2.Empirical Studies Sola and Peter (2013) examined the money supply and inflation rate in Nigerian economy. Variables that were used are inflation rate, money supply, interest rate, exchange rate, oil revenue, and government expenditure. Time series data has been used for the study purpose from 1970 to 2008. Econometric tests like Vector Auto-Regression (VAR), Augmented Dickey-Fuller, and Granger Causality were used. A positive relation between money supply and inflation rate is found also causality test shows unidirectional causality between exchange rate and inflation rate, interest rate and inflation rate, also money supply and government expenditures. Ali et al (2013) enquired the impact of government expenditures on economic growth during different democratic and dictatorial eras. The data is from the year of 1972-2009. Different sub-categories of government 1 Amedeo Strano (2002) 2 Cakrani et al (2013)

8. 8 expenditures are taken as variables. Data analysis is carried out by using Auto-Regressive Distributed Lag (ADL). It is concluded that development expenditure heavily support economic growth and it’s also validates the assumption of that public and private investments are complementary to each other. Georgantopoulos and Tsamis (2012) investigated the interrelationship of money supply, government expenditure, prices, and economic growth in both short-run and long-run in Cyprus. For this purpose the variable used are M2 (broad money), CPI, and Real GDP. Annual data is gathered form year 1980 to 2009. Error Correction Model (ECM) and Johnson co-integration techniques were used to analyze the data. Findings show that inflation effects economic growth very badly, and this is recommended to control government expenditures that increase aggregate demand and shifting of these expenditures to increase aggregate supply. Husnain et al (2011) questioned the interrelationship of government spending, foreign direct investment, and economic growth in Pakistan scenario. The data is collect from 1975 to 2008. Four variables were taken for analyses that are capital (K), labor (L), foreign direct investment (F), and total factor productivity (A). It is concluded that one most contributing factor in Pakistan’s economic growth is trade openness and the existing government expenditure structure is not in accordance to the economic needs it must be reorganized. Mehmood and Sadiq (2010) examined the effect of government expenditure in the form of fiscal deficits on level tax revenues. Using controlled variables poverty, government expenditure, remittance, private investment, and secondary school enrollment from 1976 to 2010 annually. For long-run relationship co-integration test is applied and for short-run error correction model is used. The Augmented Dickey-Fuller results show that poverty reduces due to public spending thrift and remittances. Also public spending will increase aggregate demand in long-run. Muhammad et al (2009) examined the relationship among government expenditure, M2, inflation and economic growth in Pakistan. Data was collected from the year 1977 to 2007. The method used for this purpose was Johnson co integration test to check the long term relationship, Granger causality test also has been used. The research findings revealed that there exists a negative relationship b/w public expenditure and inflation with economic growth. Also M2 is positively related to economic growth. The only reason that was determined for negative relationship by the research was due to cost push inflation. The research suggested that growth rate of money supply should be controlled. And governments must try to restrain from financing deficit by borrowing from central bank.

9. 9 Iqbal and Malik (2009) studied the relationship between government expenditure and taxes for the case of Pakistan using annual data from 1961 to 2008. The method used Vector Error Correction Model (VECM), granger, co-integration model, Augmented Dickey-Fuller and Johnson Co-integration approach. The analysis shows that in case of Pakistan budget deficit and debt have close relationship. But the budget deficits have no impact on the behavior of government taxes and expenditure in Pakistan. It means in case of Pakistan results found that taxes and spending decision have no relation and there is no long run co-integration between taxes and expenditure. Adrison (2002) explained money supply and government expenditure on economic growth in Indonesia to check the asymmetric effect. Money supply and government expenditures are used as variables to run the tests. The data collected is quarterly from 1980-1997. The technique that has been used is TARCH (Threshold Auto Regressive Conditional Heteroscedasticity). The regression model showed that the government expenditure has an asymmetric effect on economic output and also that money supply is not a contributing factor in inflation. It is recommended that reallocation of government spending more towards developmental programs. Albatel (2000) attempted to instigate the relationship between government expenditure and economic growth in Saudi Arabia. GDP, government expenditure, government consumption, private investment, oil revenues, and other revenues are variables upon which data is collected from 1964-1995. Method that is used to examine the data is Augmented Dickey-Fuller (ADF). On the basis of results it is concluded that government should continue providing economic infrastructure and social activities and on the other hand advise and encourage the private investors to play a meaningful role as well. Hsieh and Lai (1994) queried the relationship of government spending and economic growth in G-7 countries. The data is collected on Real GDP and share of government expenditure on goods and services from 1885-1997. Data analysis is done by using ADF and Phillip-PerronZt(q) test. The findings proved that relationship between government spending and economic growth can vary across time. But still government spending is the only best way possible for growth in economy.

10. 10 Chapter 3

11. 11 3. Methodology 3.1.Modeling Framework To estimate the effect of money supply (MS) and exchange rates (ER) on government expenditure (GE) in Pakistan following model is used: GE= α + β1MS +β2ER + ε Hₒ1 = MS does not affect GE Hₒ2 = ER does not affect GE Where GE is the percentage of GDP of Gross National Expenditure is taken, MS is taken as Quasi Money M2 as percentage of GDP and ER is taken as amount in rupees against 1 US dollar; Data for all research variables is from the time period of 1970 to 2012. Data for research variables under consideration is taken from World Bank3 . 3 Link address: http://data.worldbank.org/country/pakistan

12. 12 Chapter 4

13. 13 4. Estimation and Result 4.1.1. Estimate Equation (Regression) Hₒ= MS and ER does not affect GE Table 1 Variables Coefficient Prob Constant 124.0585 0.0000 MS -0.367716 0.0079 ER -0.058903 0.0045 It is shown that MS and ER have negative relation with the dependent variable GE, meaning if MS increase by 1 unit GE will decrease by 0.3167 units and if ER increase by 1 unit than GE will decrease by 0.0589 units. 4.1.2. Stationary Analysis So to check the trend in the data stationary analysis is conducted for verification. Unit root test is used for stationary analysis with Augmented Dickey-Fuller statistic. The hypothesis is that there is a trend in data or the data is non-stationary. Unit root test has been performed for all variables with level and 1st difference, and 2nd difference is also used where 1st difference shows the contradiction. The test has been run for individual intercept and individual intercept & Trend. The results have been reported in table 2. Table 2 Variables Level 1st Difference C C&T C C&T GE -2.023248 -2.148364 -6.912705 -6.822305 MS -4.407745 -4.665224 -5.342827 -6.176806 ER 2.933613 -4.665224 -3.839290 -4.988724

14. 14 4.1.3. Heteroskedasticity Test Heteroskedasticity often arises in cross sectional data but the data taken for the research purpose is already time series. The P value is 0.121 it means that there is no existence of heteroskedasticity in the data. 4.1.4. Auto Correlation LM Test When computed, the resulting number can range from 0 to 4. An autocorrelation of less than 2 represents perfect positive correlation, an increase seen in one time series will lead to a proportionate increase in the other time series, while a value of greater than 2 represents perfect negative correlation, an increase seen in one time series results in a proportionate decrease in the other time series4 . The probability value is 0 it means there exist positive autocorrelation in the data. The removal of auto correlation can be viewed in appendix. 4 Link address http://www.investopedia.com Prob 0.121 Heteroscedesticity Test Table 3 Prob 0.000 Table 4 Breusch-Godfrey Serial Correlation LMTest

15. 15 4.1.5. Stability Test CUSUM TEST CUSUM SQUARE TEST CUSUM is the difference between each measurement and the error value. We can see in the above diagram that the error limit is ±5%. All the data variable variables are under the significance level showing that data is stable. To verify the previous CUSUM SQUARE test is carried out here we can see that the provided data is crossing the significance level of 5% somewhere in year 1997 or 1998 which is not required so as to remove it we take the help of Chow’s breakpoint test. (see appendix) 4.1.6. Co-integration test Table 5 Variables Hypothesis No. of CE(s) Trace Statistics 5% Critical Values Maximum Eigen Value Statistics 5% Critical Values GE MS ER None * 32.91231 29.79707 23.72303 21.13162 At most 1 9.189281 15.49471 6.234627 14.26460 At most 2 2.954653 3.841466 2.954653 3.841466 * denotes rejection of the hypothesis at the 0.05 level It is found out that there is no co-integration between the said variables government expenditure, money supply, and exchange rates. At level “None” the p-value appeared to be 0.0212 or 2.12% less than 5% significance level. So we reject Hₒ that says variables are co-integrated. 4.1.7. Causality Analysis Table 6 Null Hypothesis F-Statistics Probability ER does not Granger Cause GE 0.43866 0.51167 GE does not Granger Cause ER 0.95095 0.33549 -20 -10 0 10 20 1975 1980 1985 1990 1995 2000 2005 2010 CUSUM 5% Significance -0.4 0.0 0.4 0.8 1.2 1.6 1975 1980 1985 1990 1995 2000 2005 2010 CUSUM of Squares 5% Significance

16. 16 MS does not Granger Cause GE 0.12182 0.72895 GE does not Granger Cause MS 0.82282 0.36993 MS does not Granger Cause ER 0.39045 0.53570 ER does not Granger Cause MS 0.00973 0.92194 It is evident from the causality test that money supply and exchange rates both combined are not causing government expenditure. But, it is the government expenditure that is causing money supply and exchange rates.                  

17. 17          Chapter 5          

18. 18 5. Conclusion  On the basis of data analysis the study is able to conclude that money supply and exchange rates are negatively affecting government expenditure. Using the annual data of the period 1970-2012, we found that there is negative relationship between money supply and exchange rates on government expenditure in long run. First of all, we have to understand that change in money supply is not the only way of financing government spending, but for past several years there has been a lot of pressure given on central bank borrowing, plus borrowing from international financial institution have rapidly change Pakistan’s economic positions. Secondly a country’s economic performance can be judged by its exchange rate values, one can realize the fact that the more rupees depreciate then more money is required by the government to pay the debts and run the economic operations. Whatever the reason is government must need to curtail its expenses and find some other solutions rather than burdening it’s on economy. Some of the policy recommendation that government authorities need to consider are a follow: 1. Government expenditure on infrastructure so as to facilitate the economic growth in the long-term scenario. 2. Developing such financial and capital markets that can mobilize savings and channel them to productive use. 3. Investment should be made in international arena in different assets with the aim of having high return rather than borrowing from World Bank or other financial bodies.   

19. 19 Bibliography

20. 20 6. References o Buchanan and Wagner 1977 “Democracy in Deficit”, New York: Academic Press o Buchanan and Wagner 1978 “Dialogues concerning Fiscal Religion”, Journal of Monetary Economics, vol.4, pp 627-36 o Blackley 1986 “Causality between Revenues and Expenditures and the Size of the Federal Budget” Public Finance Quarterly, vol.14, pp 139-56 o Ram 1988 “Additional Evidence on Causality between Government Revenue and Government Expenditure”, Southern Economic Journal vol-54, pp 763-69 o Guffey 1985 “The Federal Reserve’s Role in Promoting Economic Growth”, FRB Kansas city, Economic Review o Landau 1983 “Government Expenditure and Growth Rate: A Cross Country Study”, Southern Economic Journal, vol-49, 783-92 o Iqbal an Malik “Budget Balance through Revenue or Spending Adjustment: Evidence from Pakistan”, The Pakistan Development Review, 49:4 Part II (Winter 2010) pp. 611–630 o Jiranyakul 2007 “The Relation between Government Expenditures and Economic Growth in Thailand”, Munich Personal RePEc Archive, Paper no 46070 o Mehmood and Sadiq 2010 “The Relationship between Government Expenditure and Poverty: A Cointegration Analysis”, Romanian Journal of Fiscal Policy, Volume 1, Issue 1, July-December 2010, Pages 29-37 o Hsieh and Lai 1994 “Government spending and Economic Growth: The G-7 Experience”, California State University, Journal of Applied economics, Vol 26, 535-542 o Adrison 2002 “The Effect of Money Supply and Government Expenditure Shock in Indonesia: Symmetric or Asymmetric”, International Study Program o Strano 2002 “How and How much can the money Supply affect the Inflation Rate”,http://ukdataservice.ac.uk/media/263125/strano-paper.pdf o Sola and Peter 2013 “Money Supply and Inflation in Nigeria: Implications for National Development”, Modern Economy, http://www.scrip.org/journal/me, Vol 4, 161-170 o Cakrani et al 2013 “ Government spending and Real Exchange Rate: Case of Albania”, European Journal of Sustainable Development, Vol 2, Issue 4, 303-310 o Albatel 2000 “The Relationship between Government Expenditure and Economic Growth in Saudi Arabia”, journal of King Saud University, Vol 2, 173-191 o Georagantopoulas and Tsamis 2010 “The Interrelationship between Money Supply, Prices, Government Expenditure, and Economic Growth: A Causality Analysis for the Case of Cyprus” International Journal of Economic Science and Applied Research, Vol 5 (3), 115-128 o Mohammad et al 2009 “An Empirical Investigation between Money Supply, Government Expenditure, Output & Prices: The Pakistan Evidence”, European Journal of Economics, Finance and Administrative Sciences ISSN 1450- 2275 Issue 17

21. 21 o Husnain et al “Public Spending, Foreign Direct Investment and Economic Growth”, International Research Journal of Finance and Economics ISSN 1450-2887 Issue 61 o Ali et al 2013 “The composition of public expenditures and economic growth: evidence from Pakistan", International Journal of Social Economics, Vol. 40 Issue 11, pp.1010 – 1022

22. 22 Appendix

23. 23 APPENDIX #1 Year Exchange Rate Value Money and quasi money (M2) as % of GDP Gross national expenditure (% of GDP) 1970 4.76 43.46080808 106.9011662 1971 4.76 46.76431456 105.6476224 1972 8.68 51.30452441 105.2081712 1973 9.99 47.19722464 102.7768005 1974 9.9 35.90238629 107.096748 1975 9.9 33.66789516 111.5387994 1976 9.9 37.839395 108.652411 1977 9.9 39.35625433 109.7361804 1978 9.9 40.0350311 109.2384445 1979 9.9 43.11308147 112.0646621 1980 9.9 41.49679392 111.612632 1981 9.9 39.01479608 110.6951501 1982 11.85 40.79247985 111.8188964 1983 13.12 43.87686615 111.0590559 1984 14.05 39.85505175 111.5557487 1985 15.93 40.66128466 112.3908293 1986 16.65 43.30970533 110.7658504 1987 17.4 45.31106102 107.7687996 1988 18 41.36542167 108.0837644 1989 20.54 38.98095986 107.8638429 1990 21.71 39.13710282 107.8328811 1991 23.8 39.18989814 101.5607972 1992 25.08 42.74812397 103.169256 1993 28.11 45.65696029 106.134395 1994 30.57 45.75915506 102.7620197 1995 31.64 43.57084594 102.7128144 1996 36.08 46.04129003 104.5239233 1997 41.11 48.20323748 104.6883597 1998 45.05 47.1502314 101.0421428 1999 49.5 44.82025634 101.6129642 2000 53.65 38.59469838 101.2469581 2001 61.93 39.15125231 101.0524546 2002 59.72 43.25191223 100.0903955 2003 57.75 46.42524235 99.40656093 2004 58.26 48.36162239 98.9663297 2005 59.51 49.18651118 103.8742992 2006 60.27 44.55519367 107.413804 2007 60.74 47.43290591 106.5612017

24. 24 2008 70.41 43.54570483 110.8295797 2009 81.71 40.27346075 107.2803434 2010 85.19 41.13943884 105.8363909 2011 86.34 37.47580384 105.0042579 2012 93.4 39.9150827 107.9296846

25. 25 APPENDIX #2 Figure No 1. Figure No 2. Figure No 3. Figure No 4.

26. 26 Table 1: Regression Dependent Variable: GE Method: Least Squares Date: 12/04/14 Time: 17:30 Sample: 1970 2012 Included observations: 43 Variable Coefficient Std. Error t-Statistic Prob. ER -0.058903 0.019581 -3.008184 0.0045 MS -0.367716 0.131483 -2.796688 0.0079 C 124.0585 5.611903 22.10632 0.0000 R-squared 0.316909 Mean dependent var 106.3723 Adjusted R-squared 0.282755 S.D. dependent var 3.891667 S.E. of regression 3.295865 Akaike info criterion 5.290429 Sum squared resid 434.5090 Schwarz criterion 5.413303 Log likelihood -110.7442 F-statistic 9.278695 Durbin-Watson stat 0.593900 Prob(F-statistic) 0.000489 Table 2: Stationary analysis GE variable Level: Intercept Null Hypothesis: GE has a unit root Exogenous: Constant Lag Length: 0 (Automatic based on SIC, MAXLAG=1) t-Statistic Prob.* Augmented Dickey-Fuller test statistic -2.023248 0.2761 Test critical values: 1% level -3.596616 5% level -2.933158 10% level -2.604867 *MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller Test Equation Dependent Variable: D(GE) Method: Least Squares Date: 12/04/14 Time: 18:14 Sample (adjusted): 1971 2012 Included observations: 42 after adjustments Variable Coefficient Std. Error t-Statistic Prob. GE(-1) -0.187393 0.092620 -2.023248 0.0498 C 19.95092 9.855305 2.024384 0.0496

27. 27 R-squared 0.092837 Mean dependent var 0.024488 Adjusted R-squared 0.070158 S.D. dependent var 2.417742 S.E. of regression 2.331387 Akaike info criterion 4.577252 Sum squared resid 217.4147 Schwarz criterion 4.659998 Log likelihood -94.12229 Hannan-Quinn criter. 4.607582 F-statistic 4.093532 Durbin-Watson stat 1.998466 Prob(F-statistic) 0.049766 Level: Trend and Intercept Null Hypothesis: GE has a unit root Exogenous: Constant, Linear Trend Lag Length: 0 (Automatic - based on SIC, maxlag=1) t-Statistic Prob.* Augmented Dickey-Fuller test statistic -2.148364 0.5048 Test critical values: 1% level -4.192337 5% level -3.520787 10% level -3.191277 *MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller Test Equation Dependent Variable: D(GE) Method: Least Squares Date: 12/17/14 Time: 19:59 Sample (adjusted): 1971 2012 Included observations: 42 after adjustments Variable Coefficient Std. Error t-Statistic Prob. GE(-1) -0.226091 0.105239 -2.148364 0.0380 C 24.63691 11.55258 2.132590 0.0393 @TREND("1970") -0.026558 0.033723 -0.787544 0.4357 R-squared 0.107038 Mean dependent var 0.024488 Adjusted R-squared 0.061246 S.D. dependent var 2.417742 S.E. of regression 2.342534 Akaike info criterion 4.609093 Sum squared resid 214.0112 Schwarz criterion 4.733212 Log likelihood -93.79095 Hannan-Quinn criter. 4.654588 F-statistic 2.337446 Durbin-Watson stat 1.955270 Prob(F-statistic) 0.109961 1st Difference: Intercept Null Hypothesis: D(GE) has a unit root Exogenous: Constant Lag Length: 0 (Automatic based on SIC, MAXLAG=1) t-Statistic Prob.* Augmented Dickey-Fuller test statistic -6.912705 0.0000 Test critical values: 1% level -3.600987

28. 28 5% level -2.935001 10% level -2.605836 *MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller Test Equation Dependent Variable: D(GE,2) Method: Least Squares Date: 12/04/14 Time: 18:19 Sample (adjusted): 1972 2012 Included observations: 41 after adjustments Variable Coefficient Std. Error t-Statistic Prob. D(GE(-1)) -1.116067 0.161452 -6.912705 0.0000 C 0.050291 0.383336 0.131193 0.8963 R-squared 0.550616 Mean dependent var 0.101927 Adjusted R-squared 0.539093 S.D. dependent var 3.614790 S.E. of regression 2.454084 Akaike info criterion 4.680935 Sum squared resid 234.8786 Schwarz criterion 4.764524 Log likelihood -93.95917 Hannan-Quinn criter. 4.711373 F-statistic 47.78549 Durbin-Watson stat 2.024600 Prob(F-statistic) 0.000000 1st Difference: Trend and Intercept Null Hypothesis: D(GE) has a unit root Exogenous: Constant, Linear Trend Lag Length: 0 (Automatic - based on SIC, maxlag=1) t-Statistic Prob.* Augmented Dickey-Fuller test statistic -6.822305 0.0000 Test critical values: 1% level -4.198503 5% level -3.523623 10% level -3.192902 *MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller Test Equation Dependent Variable: D(GE,2) Method: Least Squares Date: 12/17/14 Time: 20:00 Sample (adjusted): 1972 2012 Included observations: 41 after adjustments Variable Coefficient Std. Error t-Statistic Prob. D(GE(-1)) -1.115871 0.163562 -6.822305 0.0000 C -0.011621 0.819646 -0.014178 0.9888 @TREND("1970") 0.002815 0.032815 0.085771 0.9321 R-squared 0.550703 Mean dependent var 0.101927

29. 29 Adjusted R-squared 0.527056 S.D. dependent var 3.614790 S.E. of regression 2.485924 Akaike info criterion 4.729522 Sum squared resid 234.8332 Schwarz criterion 4.854905 Log likelihood -93.95520 Hannan-Quinn criter. 4.775180 F-statistic 23.28829 Durbin-Watson stat 2.025199 Prob(F-statistic) 0.000000 MS variable Level: Intercept Null Hypothesis: MS has a unit root Exogenous: Constant Lag Length: 1 (Automatic - based on SIC, maxlag=1) t-Statistic Prob.* Augmented Dickey-Fuller test statistic -4.407745 0.0011 Test critical values: 1% level -3.600987 5% level -2.935001 10% level -2.605836 *MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller Test Equation Dependent Variable: D(MS) Method: Least Squares Date: 12/17/14 Time: 20:02 Sample (adjusted): 1972 2012 Included observations: 41 after adjustments Variable Coefficient Std. Error t-Statistic Prob. MS(-1) -0.552360 0.125316 -4.407745 0.0001 D(MS(-1)) 0.453209 0.145532 3.114156 0.0035 C 23.49588 5.380859 4.366567 0.0001 R-squared 0.355535 Mean dependent var -0.167054 Adjusted R-squared 0.321615 S.D. dependent var 3.385277 S.E. of regression 2.788253 Akaike info criterion 4.959063 Sum squared resid 295.4256 Schwarz criterion 5.084447 Log likelihood -98.66080 Hannan-Quinn criter. 5.004721 F-statistic 10.48180 Durbin-Watson stat 1.952386 Prob(F-statistic) 0.000237 Level: Trend and Intercept Null Hypothesis: MS has a unit root Exogenous: Constant, Linear Trend Lag Length: 1 (Automatic - based on SIC, maxlag=1) t-Statistic Prob.* Augmented Dickey-Fuller test statistic -4.665224 0.0029

30. 30 Test critical values: 1% level -4.198503 5% level -3.523623 10% level -3.192902 *MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller Test Equation Dependent Variable: D(MS) Method: Least Squares Date: 12/17/14 Time: 20:04 Sample (adjusted): 1972 2012 Included observations: 41 after adjustments Variable Coefficient Std. Error t-Statistic Prob. MS(-1) -0.597087 0.127987 -4.665224 0.0000 D(MS(-1)) 0.487622 0.145949 3.341049 0.0019 C 24.26573 5.346348 4.538748 0.0001 @TREND("1970") 0.052086 0.037651 1.383389 0.1748 R-squared 0.387229 Mean dependent var -0.167054 Adjusted R-squared 0.337545 S.D. dependent var 3.385277 S.E. of regression 2.755322 Akaike info criterion 4.957414 Sum squared resid 280.8967 Schwarz criterion 5.124592 Log likelihood -97.62698 Hannan-Quinn criter. 5.018291 F-statistic 7.793820 Durbin-Watson stat 2.027069 Prob(F-statistic) 0.000372 1st difference: Intercept Null Hypothesis: D(MS) has a unit root Exogenous: Constant Lag Length: 0 (Automatic based on SIC, MAXLAG=9) t-Statistic Prob.* Augmented Dickey-Fuller test statistic -5.342827 0.0001 Test critical values: 1% level -3.600987 5% level -2.935001 10% level -2.605836 *MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller Test Equation Dependent Variable: D(MS,2) Method: Least Squares Date: 12/04/14 Time: 18:23 Sample (adjusted): 1972 2012 Included observations: 41 after adjustments Variable Coefficient Std. Error t-Statistic Prob. D(MS(-1)) -0.839546 0.157135 -5.342827 0.0000 C -0.143632 0.528907 -0.271564 0.7874

31. 31 R-squared 0.422614 Mean dependent var -0.021079 Adjusted R-squared 0.407809 S.D. dependent var 4.396750 S.E. of regression 3.383472 Akaike info criterion 5.323232 Sum squared resid 446.4674 Schwarz criterion 5.406821 Log likelihood -107.1263 Hannan-Quinn criter. 5.353671 F-statistic 28.54580 Durbin-Watson stat 1.877839 Prob(F-statistic) 0.000004 1st Difference: Trend and Intercept Null Hypothesis: D(MS) has a unit root Exogenous: Constant, Linear Trend Lag Length: 1 (Automatic - based on SIC, maxlag=1) t-Statistic Prob.* Augmented Dickey-Fuller test statistic -6.176806 0.0000 Test critical values: 1% level -4.205004 5% level -3.526609 10% level -3.194611 *MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller Test Equation Dependent Variable: D(MS,2) Method: Least Squares Date: 12/17/14 Time: 20:05 Sample (adjusted): 1973 2012 Included observations: 40 after adjustments Variable Coefficient Std. Error t-Statistic Prob. D(MS(-1)) -1.191044 0.192825 -6.176806 0.0000 D(MS(-1),2) 0.393181 0.151120 2.601788 0.0134 C -0.772192 1.092379 -0.706890 0.4842 @TREND("1970") 0.022737 0.043197 0.526346 0.6019 R-squared 0.536170 Mean dependent var -0.052523 Adjusted R-squared 0.497518 S.D. dependent var 4.448090 S.E. of regression 3.153072 Akaike info criterion 5.229271 Sum squared resid 357.9071 Schwarz criterion 5.398159 Log likelihood -100.5854 Hannan-Quinn criter. 5.290335 F-statistic 13.87157 Durbin-Watson stat 1.961720 Prob(F-statistic) 0.000004 ER variable Level: Intercept Null Hypothesis: ER has a unit root Exogenous: Constant Lag Length: 0 (Automatic based on SIC, MAXLAG=1) t-Statistic Prob.*

32. 32 Augmented Dickey-Fuller test statistic 2.933613 1.0000 Test critical values: 1% level -3.596616 5% level -2.933158 10% level -2.604867 *MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller Test Equation Dependent Variable: D(ER) Method: Least Squares Date: 12/05/14 Time: 07:40 Sample (adjusted): 1971 2012 Included observations: 42 after adjustments Variable Coefficient Std. Error t-Statistic Prob. ER(-1) 0.048655 0.016585 2.933613 0.0055 C 0.531435 0.673102 0.789531 0.4345 R-squared 0.177058 Mean dependent var 2.110476 Adjusted R-squared 0.156484 S.D. dependent var 2.851876 S.E. of regression 2.619252 Akaike info criterion 4.810102 Sum squared resid 274.4191 Schwarz criterion 4.892848 Log likelihood -99.01215 F-statistic 8.606083 Durbin-Watson stat 1.445382 Prob(F-statistic) 0.005524 Level: Trend and Intercept Null Hypothesis: ER has a unit root Exogenous: Constant, Linear Trend Lag Length: 0 (Automatic - based on SIC, maxlag=1) t-Statistic Prob.* Augmented Dickey-Fuller test statistic -0.258601 0.9894 Test critical values: 1% level -4.192337 5% level -3.520787 10% level -3.191277 *MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller Test Equation Dependent Variable: D(ER) Method: Least Squares Date: 12/17/14 Time: 20:08 Sample (adjusted): 1971 2012 Included observations: 42 after adjustments

33. 33 Variable Coefficient Std. Error t-Statistic Prob. ER(-1) -0.013444 0.051989 -0.258601 0.7973 C -0.283093 0.930000 -0.304401 0.7624 @TREND("1970") 0.131623 0.104522 1.259287 0.2154 R-squared 0.209212 Mean dependent var 2.110476 Adjusted R-squared 0.168659 S.D. dependent var 2.851876 S.E. of regression 2.600280 Akaike info criterion 4.817865 Sum squared resid 263.6968 Schwarz criterion 4.941984 Log likelihood -98.17516 Hannan-Quinn criter. 4.863359 F-statistic 5.158962 Durbin-Watson stat 1.417172 Prob(F-statistic) 0.010284 1st Difference: Intercept Null Hypothesis: D(ER) has a unit root Exogenous: Constant Lag Length: 0 (Automatic based on SIC, MAXLAG=1) t-Statistic Prob.* Augmented Dickey-Fuller test statistic -3.839290 0.0053 Test critical values: 1% level -3.600987 5% level -2.935001 10% level -2.605836 *MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller Test Equation Dependent Variable: D(ER,2) Method: Least Squares Date: 12/05/14 Time: 07:42 Sample (adjusted): 1972 2012 Included observations: 41 after adjustments Variable Coefficient Std. Error t-Statistic Prob. D(ER(-1)) -0.580057 0.151084 -3.839290 0.0004 C 1.326366 0.511912 2.591006 0.0134 R-squared 0.274286 Mean dependent var 0.172195 Adjusted R-squared 0.255678 S.D. dependent var 3.075190 S.E. of regression 2.653093 Akaike info criterion 4.836880 Sum squared resid 274.5172 Schwarz criterion 4.920469 Log likelihood -97.15604 F-statistic 14.74015 Durbin-Watson stat 1.733545 Prob(F-statistic) 0.000441

34. 34 1st Difference: Trend and Intercept Null Hypothesis: D(ER) has a unit root Exogenous: Constant, Linear Trend Lag Length: 1 (Automatic - based on SIC, maxlag=1) t-Statistic Prob.* Augmented Dickey-Fuller test statistic -4.988724 0.0012 Test critical values: 1% level -4.205004 5% level -3.526609 10% level -3.194611 *MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller Test Equation Dependent Variable: D(ER,2) Method: Least Squares Date: 12/17/14 Time: 20:10 Sample (adjusted): 1973 2012 Included observations: 40 after adjustments Variable Coefficient Std. Error t-Statistic Prob. D(ER(-1)) -0.930109 0.186442 -4.988724 0.0000 D(ER(-1),2) 0.303859 0.160752 1.890238 0.0668 C -0.683185 0.832711 -0.820435 0.4174 @TREND("1970") 0.117774 0.037899 3.107564 0.0037 R-squared 0.431053 Mean dependent var 0.078500 Adjusted R-squared 0.383641 S.D. dependent var 3.054524 S.E. of regression 2.398062 Akaike info criterion 4.681838 Sum squared resid 207.0252 Schwarz criterion 4.850726 Log likelihood -89.63676 Hannan-Quinn criter. 4.742903 F-statistic 9.091606 Durbin-Watson stat 1.864814 Prob(F-statistic) 0.000129 Table 3 Heteroscedesticity Test White Heteroskedasticity Test: F-statistic 1.884473 Probability 0.120651 Obs*R-squared 8.727726 Probability 0.120430 Test Equation: Dependent Variable: RESID^2 Method: Least Squares Date: 12/04/14 Time: 17:35 Sample: 1970 2012 Included observations: 43

35. 35 Variable Coefficient Std. Error t-Statistic Prob. C -228.0955 168.4667 -1.353950 0.1840 ER 0.810815 1.084077 0.747931 0.4592 ER^2 -0.005374 0.003680 -1.460195 0.1527 ER*MS -0.005402 0.020903 -0.258439 0.7975 MS 11.18316 7.991880 1.399315 0.1700 MS^2 -0.135581 0.094344 -1.437096 0.1591 R-squared 0.202970 Mean dependent var 10.10486 Adjusted R-squared 0.095264 S.D. dependent var 11.17521 S.E. of regression 10.62959 Akaike info criterion 7.693948 Sum squared resid 4180.563 Schwarz criterion 7.939697 Log likelihood -159.4199 F-statistic 1.884473 Durbin-Watson stat 1.824780 Prob(F-statistic) 0.120651 Table 4 Auto-Correlation LM Test Breusch-Godfrey Serial Correlation LM Test: F-statistic 37.80496 Probability 0.000000 Obs*R-squared 21.16547 Probability 0.000004 Test Equation: Dependent Variable: RESID Method: Least Squares Date: 12/04/14 Time: 17:37 Presample missing value lagged residuals set to zero. Variable Coefficient Std. Error t-Statistic Prob. ER 0.003985 0.014146 0.281704 0.7797 MS 0.060006 0.095387 0.629077 0.5330 C -2.627672 4.072397 -0.645240 0.5226 RESID(-1) 0.720410 0.117167 6.148574 0.0000 R-squared 0.492220 Mean dependent var -7.44E-15 Adjusted R-squared 0.453160 S.D. dependent var 3.216435 S.E. of regression 2.378509 Akaike info criterion 4.659233 Sum squared resid 220.6349 Schwarz criterion 4.823065 Log likelihood -96.17350 F-statistic 12.60165 Durbin-Watson stat 1.972120 Prob(F-statistic) 0.000007

36. 36 Chow Breakpoint Test Chow Breakpoint Test: 1998 F-statistic 5.693700 Probability 0.002614 Log likelihood ratio 16.32138 Probability 0.000974 Chow Breakpoint Test: 1999 F-statistic 5.655231 Probability 0.002713 Log likelihood ratio 16.22952 Probability 0.001017 Table 5 Co-integration Test Date: 12/17/14 Time: 21:19 Sample (adjusted): 1972 2012 Included observations: 41 after adjustments Trend assumption: Linear deterministic trend Series: GE MS ER Lags interval (in first differences): 1 to 1 Unrestricted Cointegration Rank Test (Trace) Hypothesized Trace 0.05 No. of CE(s) Eigenvalue Statistic Critical Value Prob.** None * 0.439323 32.91231 29.79707 0.0212 At most 1 0.141067 9.189281 15.49471 0.3482 At most 2 0.069529 2.954653 3.841466 0.0856 Trace test indicates 1 cointegratingeqn(s) at the 0.05 level * denotes rejection of the hypothesis at the 0.05 level **MacKinnon-Haug-Michelis (1999) p-values Unrestricted Cointegration Rank Test (Maximum Eigenvalue) Hypothesized Max-Eigen 0.05 No. of CE(s) Eigenvalue Statistic Critical Value Prob.** None * 0.439323 23.72303 21.13162 0.0211 At most 1 0.141067 6.234627 14.26460 0.5833 At most 2 0.069529 2.954653 3.841466 0.0856 Max-eigenvalue test indicates 1 cointegratingeqn(s) at the 0.05 level * denotes rejection of the hypothesis at the 0.05 level **MacKinnon-Haug-Michelis (1999) p-values Unrestricted Cointegrating Coefficients (normalized by b'*S11*b=I): GE MS ER 0.117424 0.323591 0.006315

37. 37 -0.170243 -0.040172 0.025722 -0.279652 -0.009059 -0.047993 Unrestricted Adjustment Coefficients (alpha): D(GE) 0.402600 0.686293 0.254698 D(MS) -2.033177 0.279529 -0.233012 D(ER) 0.686581 0.429384 -0.537399 1 Cointegrating Equation(s): Log likelihood -280.1015 Normalized cointegrating coefficients (standard error in parentheses) GE MS ER 1.000000 2.755755 0.053777 (0.47528) (0.07544) Adjustment coefficients (standard error in parentheses) D(GE) 0.047275 (0.04142) D(MS) -0.238743 (0.05032) D(ER) 0.080621 (0.04819) 2 Cointegrating Equation(s): Log likelihood -276.9842 Normalized cointegrating coefficients (standard error in parentheses) GE MS ER 1.000000 0.000000 -0.170274 (0.12266) 0.000000 1.000000 0.081303 (0.05334) Adjustment coefficients (standard error in parentheses) D(GE) -0.069561 0.102708 (0.06900) (0.10880) D(MS) -0.286331 -0.669147 (0.08810) (0.13890) D(ER) 0.007521 0.204922 (0.08357) (0.13176)

38. 38 Table 6 Granger’s Causality Test Pairwise Granger Causality Tests Date: 12/04/14 Time: 17:56 Sample: 1970 2012 Lags: 1 Null Hypothesis: Obs F-Statistic Probability ER does not Granger Cause GE 42 0.43866 0.51167 GE does not Granger Cause ER 0.95095 0.33549 MS does not Granger Cause GE 42 0.12182 0.72895 GE does not Granger Cause MS 0.82282 0.36993 MS does not Granger Cause ER 42 0.39045 0.53570 ER does not Granger Cause MS 0.00973 0.92194

Add a comment

Related pages

The dynamics of money supply, exchange rate and inflation ...

between exchange rate, money supply and ... financing or government expenditure, ... 138 The dynamics of money supply, exchange rate and inflation in ...
Read more

Money supply - Wikipedia, the free encyclopedia

There is strong empirical evidence of a direct relation between money ... In the money supply statistics, central bank money is MB ... the exchange rate is ...
Read more

Macroeconomics: Currency | Investopedia

Macroeconomics: Currency. Share ... Government - Expenditures, ... Relative interest rates play a major role in exchange rates between countries.
Read more

Effect of Monetary-Fiscal Policies Interaction on Price ...

The data for money supply (broad money M2), exchange rate and ... Effect of Monetary-Fiscal Policies Interaction ... Government Expenditure, Money Supply ...
Read more

Currency Converter | Foreign Exchange Rates | OANDA

Currency Converter. OANDA's currency calculator tools use OANDA Rates ™, the touchstone foreign exchange rates compiled from leading market data ...
Read more

Journal of Economic Cooperation and Development, 33, 3 ...

as money supply, prices, interest rate, exchange rate ... Output in Pakistan interrelationship among money, ... Journal of Economic Cooperation and ...
Read more

Shadowstats.com - Shadow Government Statistics - Home Page

Shadow Government Statistics Analysis Behind and Beyond Government ... Broad Money Supply ... ShadowStats.com (Shadow Government Statistics) ...
Read more

Statistics and Databases | Asian Development Bank

Asian Development Bank. ... and retail sales for the real sector; trade, exchange rates, ... and money supply, stock prices, ...
Read more

Demand for money - Wikipedia, the free encyclopedia

... stressed the choice between money ... which says that the growth rate of money supply plus the growth ... The Demand for Money: Theories, Evidence, ...
Read more