The Relationship between Macroeconomic Variables and Foreign Exchange Rate: Evidence from a Sample of Emerging (Brazil & South Africa) and Advanced (Germany & the United Kingdom) Economies By Soutonnoma Ouedraogo A research project submitted in partial fulfilment of the requirements for the degree of Master of Science in Applied Economic, College of Business and Economics. At The University of Wisconsin Whitewater July, 2018
40
Embed
The Relationship between Macroeconomic Variables and ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
The Relationship between Macroeconomic Variables and Foreign Exchange Rate:
Evidence from a Sample of Emerging (Brazil & South Africa) and Advanced
(Germany & the United Kingdom) Economies
By
Soutonnoma Ouedraogo
A research project submitted in partial fulfilment of the requirements for the degree of Master of
Science in Applied Economic, College of Business and Economics.
At
The University of Wisconsin Whitewater
July, 2018
Graduate Studies
The members of the committee approve the thesis of Soutonnoma Ouedraogo presented on August 10th, 2018.
Dr. Ahmad Yamin, Committee Chair
Dr. Ersal Eylem, Committee Member
Dr. Guo Nick, Committee Member
iii
Dedication
First and foremost, I would like to thank the almighty God for granting me the needed strength and knowledge that helped me to carry out this research.
I am also grateful to my parents; Geoffrey W. and Kathleen Thompson, Rachel Azanleko-Akouete and to my friends Kanazoe Mouni, Kanga Patrice, Sawadogo Saidou, Koubem Sibiri, Zoungrana Blanche, and Mag Fowe for being so patient with me during my Master studies. Thank you for supporting my lifelong dream.
I would also like to send thanks to my brothers (Patrick, Julien, Joel, Thierry, and Etienne Parfait) for their everyday love and care that they showed on me during my Master Studies.
Finally, I would like to thank and dedicate this thesis to my mother, Ms. Sibidou Marie Claire Kabore. It was you who originally generated my love for education since the first day you walked with me to class. Although it has been years since you have passed, I still take your advice with me, every day.
iv
Acknowledgements
I would like to express my sincere gratitude to my former advisor at the University of Wisconsin-Madison economic program, Ms. Susan C. Herring, for all the support.
I would like to express my gratefulness to the committee members who provided me the prompt guidance, encouragements, and feedback for the completion of this Master thesis. Your intellectual directions and suggestions assisted me to improve the quality of the thesis.
To Dr. Ahmad Yamin, (Committee Chair), Dr. Ersal Eylem (Committee Member), and Dr. Guo Nick (Committee Member), I am grateful for your support.
v
The Relationship between Macroeconomic Variables and Foreign Exchange Rate:
Evidence from a Sample of Emerging (Brazil & South Africa) and Advanced
(Germany & the United Kingdom) Economies
By
Soutonnoma Ouedraogo
The University of Whitewater-Wisconsin, 2018. Under the supervision of Dr. Ahmad Yamin.
Abstract
This study tries to determine the long-run relationship between macroeconomic variables such as interest rates, inflation, money supply, productivities, trade balances and exchange rate of Brazil, South Africa, Germany, and the United Kingdom. The study uses monthly data from January 1999 to December 2016. I use co-integration methods and Granger causality tests to examine the impact of the macroeconomic indicators on the exchange rate in the long-run. I find that in the long-run the differentials for interest rates, inflation, money supply, and trade balance have a significant adverse effect and the productivity differential has a significant positive impact on the Real/Dollar rate for Brazil. The study also reveals a significant negative relationship between money supply differential and the Rand/Dollar rate for South Africa, while trade balance has a significant positive effect on the Rand/Dollar exchange rate in the long-run. I find that trade balance has a negative significant impact on the Euro/Dollar in the long-run for Germany. In the long-run differentials for inflation and productivity have a significant negative impact, but differentials for the interest rates, money supply, and trade balance have a significant positive effect on the Pound/Dollar for the United Kingdom. Lastly, I find that the differentials for interest rates for Brazil and money supply for South Africa, significantly help to predict the future level of their exchange rates in the short-run. However, I do not find any evidence of a short-run relationship between the selected macroeconomic variables and the exchange rates for both Germany and the United Kingdom.
The coefficients show how deviations from the long-run relationship affect the changes in the
variables in the following period.
The results display in table 7 for Brazil show that the differentials for interest rate, inflation,
money supply, and the trade balance have a negative relationship with the exchange rate and
statistically significant at 5% level. On the other hand, productivity has a positive relationship with
the exchange rate and is statistically significant. We can interpret the coefficients as follows: For
example, a 1% increase in the home country interest rate with all the other variables being constant
(inflation, money supply, productivity, and trade balance), leads to a 0.86% depreciation on the
Real (Brazil's currency) against the US Dollar in the long-run. If Brazil has 1 unit increase in
productivity with all other variables being constant (differentials for interest rate, inflation, money
supply, and trade balance), it will lead to a 0.13% appreciation on the Real (Brazil's currency)
against the US Dollar in the long-run.
For South Africa, table 7 results show that differential for money supply has a negative
relationship with the exchange rate and statistically significant at 5% level. This implies that if
South Africa increases its money supply by one billion Rand, it will lead to less than 0.07%
depreciation of its currency against the US Dollar exchange rate. On the other hand, differentials
for interest rate, inflation, productivity, and trade balance have a positive relationship with the
exchange rate, although only trade balance's t-value indicates significance at the 5% level in the
co-integration equation. It implies that an increase of South African exportations toward the United
States by one million dollars, holding other variables constant leads to a 6.5% appreciation of their
exchange rate in the long-run.
Table 7 shows that for Germany, differentials for inflation, money supply, productivity,
and trade balance have a negative relationship with the exchange rate and only trade balance is
statistically significant at 5% level. In contrast, the interest rate differential has a positive
14
relationship with the exchange rate, but not statistically significant. Since Germany is part of the
European Union, a change in any of these variables should not be statistically significant because
the major decisions that can impact the Euro/Dollar exchange are mostly adopted as a group (The
European Union) not individually.
The results display from table 7 show that for the United Kingdom, differentials for
inflation and productivity have a negative relationship with the exchange rate and their coefficients
are statistically significant at 5% level. It indicates that 1 unit increase in the consumer price index
in the United Kingdom holding all other variables constant, leads to a 145.28% depreciation of the
Pound/Dollar exchange rate in long-run. Conversely, differentials for interest rate, money supply,
and trade balance are statistically significant at 5% level and have a positive relationship with the
exchange rate. For example, an increase of 1% short-term interest rates (T-bills) in the United
Kingdom holding other variables constant, leads to a 3.24% appreciation of Pound/Dollar
exchange rate in the long-run. Here we can see that the common statement (an increase of interest
rate and inflation in the home country brings appreciation and depreciation of exchange rate) of
relative interest rates and inflation is satisfied.
Error Correction Estimates
From table 8 in the appendix, the speed of adjustment back to the equilibrium is represented
by the error correction estimates. The value of error correction term should lie between 0 and -1.
A negative sign indicates the convergence and evaluates the speed of adjustment toward the
equilibrium. It can be seen that for Brazil, four variables (exchange rate, differentials for money
supply, productivity, and trade balance) have negative error correction terms, except the exchange
rate and differential productivity's values lines within range 0 and -1. Also, exchange rate only is
statistically significant at 5% level. It implies that a 1.40% of disequilibrium was corrected each
month by changes in exchange rate. For South Africa, only inflation differential has a negative
sign value which lies between 0 and -1, and statistically significant at 5% level. The results from
table 8 indicate that 0.67% disequilibrium was corrected each month by changes in relative prices.
For Germany, exchange rate, differentials for money supply, productivity, and trade
balance have a negative sing of error correction terms. However, three variables (exchange rates,
differentials for money supply and productivity) values line between 0 and -1. And the exchange
rate alone is statistically significant at 5% level. Table 8 shows that for the United Kingdom have
15
four variables (exchange rate, differentials for interest rates, inflation, and productivity) have
negative error correction terms with all them lining within 0 and -1. Conversely, both interest rates
and inflation differentials are statistically significant at 5% level.
Granger Causality
VEC Granger-causality test was taken into account due to the existence of the co-
integration relationship in order to determine which variable better helps in forecasting the
exchange rates in short-run. VEC Granger-causality can accommodate more than two variables
while the pairwise is meant for two variables at a time and short-run. Looking at the results in table
9 in the appendix for Brazil, only the Wald tests statistic (0.024) for the interest rates differential
is statistically significant at 5% level in exchange rate equation. It suggests that there is short-run
causal relationship running from interest rate differential to the exchange rate. In order words,
interest rates differential contributes considerably to changes in the exchange rate, which would
lend support to the interest parity idea. Moreover, the study finds that when all variables are taken
together, Wald statistic (0.09) is insignificant at 5% level. It indicates that in short-run all variables,
when are taken together, do not contribute to the changes in Real/US-dollar exchange rate changes.
For South Africa, only Wald statistic (0.043) for money supply differential is significant
at 5% level in exchange rate equation. It means that differential for money supply has an effect on
the exchange rate in short-run and that causal relationship runs from money supply differential to
the exchange rate. Therefore, differential for money supply contributes significantly to the changes
of the exchange rate between South Africa's Rand and the United States Dollar. When all variables
are taken together, the Wald statistic (0.20) is insignificant and means that as a group the variables
do not contribute considerably to the changes of the Rand/Dollar exchange rate.
It can be seen from the results in table 9 that for Germany, none of the variables' Wald
statistic is statistically significant. Similarly, when all variables are taken together, the Wald
statistic is still insignificant. This suggests that the selected macroeconomic indicators for
Germany do not contribute to the changes of the Euro/Dollar exchange rate. In other words, they
do not have an impact on Germany-US exchange rate. Likewise, at 5% significance level, none of
the variables' Wald statistic is significant for the United Kingdom. Also, all taken together, the
Wald statistic (0.39) is insignificant at 5% level. These results mean that individually and
collectively, the variables do not influence the Pound/Dollar exchange rate in short-run.
16
Impulse Response Function
Figure 1 plots the response of the Real/Dollar exchange rate to a shock in differentials for
interest rates, inflation, money supply, productivity, and trade balance. The response is not
significantly different from zero. The most interesting result from the figure 1 is the responses of
the exchange rate to a shock in interest rate differential and inflation differential. The exchange
rate decreases whenever there is one standard deviation positive shock in the interest rate.
Similarly, exchange rate falls when there is a positive shock in inflation. It implies that there is a
negative relationship between exchange rate and variables such as interest rate and inflation. The
only one variable that seems to be a significant response is from the money supply differential.
When, the response is insignificant it means that there is not strong statistical evidence that the
response is different from zero. The response of exchange rate to the selected macroeconomic
variables for South Africa is plotted in figure 2. It can be seen that there is a positive relationship
between exchange rate and interest rates differential as whenever differential for interest rates rises
exchange rate does increase too. This is not the same with the differentials for inflation and money
supply. It shows that those two variables have a negative relationship with the exchange rate, but
the money supply differential is the only one that is significant in the long run, which supports the
VECM results.
Figure 3 shows that for Germany there is a negative relationship between exchange rate
and interest rate differential. Exchange rate responds negatively to a positive one standard
deviation shock in the interest rates differential. Meanwhile, differential for inflation displays
positive association with the exchange rate. However, none of the responses are statistically
significant. For the United Kingdom, figure 4 shows that there is an association between exchange
rate and interest rates differential from period one to period 6 and then the relationship becomes
adverse to the 12th period. Also, it clearly shows whenever there is one standard deviation shock
of differentials for inflation or productivity or trade balance, exchange rate responds negatively to
their shocks. Similar as Germany, none of the responses are statistically significant. This indicates
the inability of the exchange rates of both countries to capture any short-run relationship. In other
words, the short-run relationship is being wipe away.
17
Variance Decomposition
The first part of table 9 in the appendix shows that for Brazil the variance decomposition
of exchange rate following a shock to exchange rate innovations of 0.09%. In the 1st period, the
entire change in exchange rate is explained only by a shock to exchange rate innovation. This
shock also causes immediate changes in differentials for interest rates, inflation, money supply,
productivity, and trade balance in the 2nd period. It can be seen that differentials for interest rates,
inflation, money supply, productivity, and trade balance variables account for 0.13%, 0.60%,
0.52%, 0.5144%, and less than 0.05% of the change, respectively in the exchange rate. In this
period we can see that inflation and money supply have the highest impact on exchange rate. An
excess of currency in circulation impacts prices and therefore influences the exchange rate. When
the entire 12 periods is taken into account, money supply and trade balance have the highest (3.33%
and 4.23%) total variations in exchange rate.
For the second part of table 9, which traces the variance decomposition of the exchange
rate for South Africa, presents the reactions following a shock to exchange rate innovations of
0.33%. The influence of interest rates differential on exchange rate increases over time, and by the
end of the 12th period, it is accounting for 2.49% of the total variation in exchange rate. Similarly,
differential for money supply accounts for the highest total variation of 5.22% in the exchange
rate. Also, fluctuation in money supply differential in circulation affects the Rand/Dollar exchange
rate.
Table 10 in the appendix shows that for Germany, the variance decomposition of exchange
rate following a shock to the exchange rate innovations of 0.03%. In the 2nd period, a shock to the
exchange rate innovations accounts for almost the entire change in exchange rate variation
(99.85%). This situation is nearly the same for all periods, for example, at the end of the 12th
period, the change in the exchange rate because of the initial shock is accounted by the exchange
rate itself is 99.67%. Over the entire 12 periods, none of the independent variables accounts an
impact more than 0.5% on the exchange rate. The results indicate that for Germany, the selected
macroeconomic variables do not contribute a lot in the movements of the Euro/Dollar exchange
rate. That can be the fact that Germany is in the European Union and alone does not have a
substantial effect.
18
From table 10, it can be seen that for the United Kingdom the variance decomposition of
exchange rate presents the reactions following a shock to exchange rate innovations of 0.03%. The
influence of interest rate on exchange rate fluctuations (increases and decreases) in time, and by
the end of the 12th period, it is accounting for 0.30% of the total variation in exchange rate. On the
contrary, trade balance on exchange rate increase over the time and accounts the highest total
variation of 6.12% in the exchange rate in the 12th period. The exchange rate accounts for 91.98%
of its own change in the 12th period. Here it is clear that having the US as the most significant
trading partner, the Pound/Dollar exchange rate could suffer due to a fluctuation of the trade
balance between the United Kingdom and the United States.
Conclusion
This study is designed to explore the causal relationship between the foreign exchange rates
and the macroeconomic variables such as differentials for interest rate, inflation, money supply,
productivity, and trade balance. Monthly data from 1999 to 2016 have been used and various
econometrics tests including the unit root, co-integration, vector error correction mechanism,
Granger-causality, impulse response functions, and variance decomposition are being conducted.
These analyses are useful to gauge long and short-run relationships, causal effects, and the shock
responses among the variables within a system.
The empirical evidence indicates that there exist more than one co-integrating vector for
all equations and this indicates which suitable methodology to adopt. Through the evidence, for
emerging countries the results show that in the long-run the differentials for interest rates, inflation,
money supply, and trade balance have an adverse significant impact on the Brazil/US exchange
rate. It implies that the exchange rate depreciates as the differentials for interest rates, inflation,
money supply, and trade balance decrease. Conversely, productivity differential has a significant
positive impact on the Brazil/US exchange rate, which means that an appreciation of the foreign
exchange rate is associated with an increase of home country’s total industrial productivity.
Increasing interest rates differential leads to an increase in investments which raises output, price
level increases. Therefore the value of the currency goes down. The findings show that for Brazil,
the current data of interest rates differential help to predict the future level of the Real/Dollar
exchange rate in the short-run. These findings are different to what has been found in the case of
a country like South Africa. The study reveals a significant negative relationship between money
19
supply differential and the Rand/Dollar exchange rate in the long-rum for South Africa. On the
other hand, differentials for interest rates, inflation, productivity, and trade balance have a positive
effect on the Rand/Dollar exchange rate, but only trade balance's effect is statistically significant.
For short-run, the empirical evidence show that current data of differential for money can help in
forecasting the future level of the Rand/Dollar exchange rate.
On the other hand, for advanced economies the study presents evidence that in the long-
run differentials for inflation, money supply, productivity, and trade balance have negative
influence on the Euro/Dollar for Germany. However, only trade balance has a significant long-run
impact. The evidence also shows that interest rates differential has a positive insignificant
relationship with the Euro/Dollar in the long-run. The results indicate that differentials for inflation
and productivity have a significant adverse impact on the Pound/Dollar exchange rate. In the
meantime, differentials for interest rates, money supply, and trade balance have a positive
significant effect on the Pound/Dollar exchange rate for the United Kingdom. It implies that an
increase of interest rate and inflation brings respectively an appreciation and a depreciation of the
Pound/Dollar exchange rate. Lastly, there is no evidence that the current data of any of the selected
macroeconomic variables help to predict the future level of the exchange rates for both Germany
and the United Kingdom.
From the policy perspectives, the study recommends that policy makers in Brazil and South
Africa should be cautious about the positive and negative effects of total industrial productivity
and money supply respectively on their exchange rates against the US Dollar. Contrary the
evidence found in this study indicate that the negative effects of the differentials for inflation and
total industrial productivity on the exchange rate are not negligible for both of the Advanced
countries. Since the study was based on monthly data, it could have some limitations compared to
quarterly or yearly date which could give more precise results. Also, the study did not consider all
of the macroeconomics variables affecting the exchange rate in these countries. Thus, this study
recommends that further studies be conducted to incorporate variables such as Gross Domestic
Product (GDP), unemployment rates, current accounts, public debt, Foreign Direct Investment,
etc.
20
References
Literature
Alex .H. and Innes .M. (2006). The relationship between interest rates and inflation in South Africa. Revisiting Fishers' Hypothesis. Euro Journals Publishers Inc.
Bahmani-Oskooee, Mohsen and A.B.M. Nasir (2004). “ARDL Approach to Test the Productivity Bias Hypothesis. Review of Development Economics, 8(3), 483-88.
Bailey, A., & Millard, S. (2001). Capital flows and exchange rates. Bank of England Quarterly Bulletin, Autumn.
Bergen, J. (2010). 6 Factors that influence exchange rates. Retrieved from Investopedia:http://www.investopedia.com/articles/basics/04/050704.asp
Bhundia, A., and J. Gollschalk (2003), “Sources of National Exchange Rate Fluctuations in South Africa,” IMF Working paper, WP/03/252 (December, 2003).
Blundell-Wignall, A and Browne, F (1991), ‘Increasing financial market integration: real exchange rates and macroeconomic adjustment,' OECD working paper.
Burda, M., Wyplosz, C. (2009). Macroeconomics: a European text. 5th ed. Oxford: Oxford University Press.
Campbell, J.Y. and R.H. Clarida, 1987, The dollar and real interest rates, Carnegie-Rochester Conference Series on Public Policy 27, 103-140.
Chen, Y. C., & Rogoff, K. (2003). Commodity currencies. Journal of International Economics, 60(1), 133-160.
Cheung, Y. W., Chinn, M. D., & Pascual, A. G. (2005). Empirical exchange rate models of the nineties: Are any fit to survive? Journal of International Money and Finance, 24(7), 1150 1175.
Chinn, M. and L. Johnson (1999), "The impact of Productivity Differentials on Real Exchange Rates: Beyond the Balassa Samuelson Framework," University of California Santa Cruz, Department of Economics, Working Paper No. 442.
Coughlin, C C and Koedijk, K (1990), ‘What do we know about the long-run real exchange rate?' Federal Reserve Bank of St Louis Review, January/February, pages 36-48.
Dickey, D. A., & Fuller, W. A. (1981 July). Likelihood ratio statistics for autoregressive time series with a unit root. Econometrica, 49(4), 1057–1072.
Dickey, D.A, Fuller, W.A. (1979) "Distributions of the Estimators for Autoregressive TimeSeries with a Unit Root", Journal of American Statistical Association 74:366, pp. 427-481.
21
Dong, W., & Nam, D. (2011). Exchange rates and individual good's price misalignment: Some preliminary evidence of long-horizon predictability (Discussion Paper No. 2011-8). Bank of Canada.
Dornbusch, R. (1976), “Devaluation, Money and Non-TRADED Goods,” American Economic Review, 63(5), 871-880.
Duarte, Margarida & Stockman, Alan C., 2002. "Comment on Exchange rate pass-through, exchange rate volatility, and exchange rate disconnect," Journal of Monetary Economics, Elsevier, vol. 49(5), pages 941-946, July.
Edison, J.H. and B.D. Pauls, 1993, A re-assessment of the relationship between real exchange rates and real interest rates: 1974-1990, Journal of Monetary Economics 31, 14991-64.
Engle, R.F. and C.W.J. Granger, 1987. Co-integration and error-correction: Representation, estimation and testing. Econometrica, 55(2): 251-276.
Ezirim, C. B., Edith AzukaAmuzie and Michael I. Muoghalu (2012). Autoregressive Distributed Lag Analysis of Interdependencies Between Inflation and Exchange Rates in Sub-Saharan Nigeria. The IABPAD Conference Proceedings Dallas, Texas, Vol. 9, No. 2, April 19-22, 2012, pp. 1082-1093.
FISHER, I., 1930. The Theory of Interest. New York: Macmillan.
Hushmand, M., daneshnia, M., Shahrivar, S., Qizilbash, A. and Eskandari Pour, Z. (2012). "Relationship between monetary policy and exchange rate in Iran.‟ Journal of Quantitative Economics, 272: 109-1.
Johansen, S. (1991). Estimation and Hypothesis Testing of Cointegrating Vectors in Gaussian Vector Autoregressive Models. Econometrica, Vol. 59, pp. 1551–1580.
Johansen, S., & Juselius, K. (1990). Maximum likelihood estimation and inference on cointegration - with applications to the demand for money. Oxford Bulletin of Economics and Statistics, 52 (2), 169-210.
Jonathan Kearns & Phil Manners, 2006. "The Impact of Monetary Policy on the Exchange Rate: A Study Using Intraday Data," International Journal of Central Banking, International Journal of Central Banking, vol. 2(4), December.
Keynes, J.M. (1923). A tract on monetary reform. Macmillan and St. Martin's Press for the Royal Economics Society, London.
Khan, George A. 1989. "The Output and Inflation Effects of Dollar Depreciation." Federal Reserve Bank of Kansas City, Research Working Paper 85-05. Kansas City: Federal Reserve Bank of Kansas City.
Lee, Jaewoo and Man-Keung Tang (2007). “Does Productivity Growth Appreciate the Real Exchange Rtes? Review of International Economics, 15(1), 164-87.
22
Madura, J., 2000. International financial management. 6 edition, South-Western College Publishing.
Onafowora, O. (2003). Exchange rate and trade balance in East Asia: Is there a J-curve. Economic Bulletin, Vol. 5, No.18, pp. 1-13.
Phillips, P. and Perron, P. (1988), "Testing for a Unit Root in Time Series Regression," Biometrica 75(2), pp. 335-346.
Rose, A.K., & Yellen, J.L. (1989). Is there a J-curve. Journal of Monetary Economics, 24, 53-68.
Stockman, A. C. (1980). A theory of exchange rate determination. The Journal of Political Economy, 84(4), 673-698.
Data
Variables Source Foreign Exchange rates Real/dollar https://fred.stlouisfed.org/series/EXBZUS Euro/dollar https://fred.stlouisfed.org/series/EXUSEU Rand/dollar https://fred.stlouisfed.org/series/EXSFUS Pound/dollar https://fred.stlouisfed.org/series/EXUSUK Interest Rates Brazil https://fred.stlouisfed.org/series/INTGSTBRM193N Germany https://fred.stlouisfed.org/series/IR3TIB01DEM156N South Africa https://fred.stlouisfed.org/series/IR3TTS01ZAM156N United Kingdom https://fred.stlouisfed.org/series/IR3TTS01GBM156N United States https://fred.stlouisfed.org/series/TB3MS Inflation Brazil https://fred.stlouisfed.org/series/BRACPIALLMINMEI Germany https://fred.stlouisfed.org/series/CPHPTT01DEM659N South Africa https://fred.stlouisfed.org/series/ZAFCPIALLMINMEI United Kingdom https://fred.stlouisfed.org/series/GBRCPIALLMINMEI United States https://fred.stlouisfed.org/series/GBRCPIALLMINMEI Money Supply Brazil https://fred.stlouisfed.org/series/MYAGM1BRM189N Germany https://fred.stlouisfed.org/series/MYAGM1EZM196N South Africa https://fred.stlouisfed.org/series/MAM1A1ZAM189N United Kingdom https://fred.stlouisfed.org/series/MANMM101GBM189S United States https://fred.stlouisfed.org/series/MYAGM1USM052S Productivity Brazil https://fred.stlouisfed.org/series/BRAPROINDMISMEI Germany https://fred.stlouisfed.org/series/ZAFPROMANMISMEI South Africa https://fred.stlouisfed.org/series/ZAFPROMANMISMEI United Kingdom https://fred.stlouisfed.org/series/GBRPROINDMISMEI United States https://fred.stlouisfed.org/series/USAPROINDMISMEI Trade balance
Brazil https://www.census.gov/foreign-trade/balance/c3510.html Germany https://www.census.gov/foreign-trade/balance/c4280.html South Africa https://www.census.gov/foreign-trade/balance/c7910.html United Kingdom https://www.census.gov/foreign-trade/balance/c4120.html
Appendix
Test-statistic p-value Adj. Test-statistic p-valueEmerging Countries
Notes: (1) The equations include constant and linear trend. Null hypothesis: Series has unit root(2) *, **, *** denote rejection of null hypothesis at the 10, 5, 1% level of significance, respectively
Table1: ADF and PP unit root tests on level series: Advanced and Emerging individual economies
Notes: (1) The equations include constant and linear trend. Null hypothesis: Series has unit root(2) *, **, *** denote rejection of null hypothesis at the 10, 5, 1% level of significance, respectively
Table2: ADF and PP unit root tests on first-difference series: Advanced and Emerging individual economies
ADF PP
𝚤̃𝜋�𝑚�𝑦𝑦�
𝑄𝑡
𝑇𝑏
𝚤̃𝜋�𝑚�𝑦𝑦�
𝑄𝑡
𝑇𝑏
𝚤̃𝜋�𝑚�𝑦𝑦�
𝑄𝑡
𝑇𝑏
𝚤̃𝜋�𝑚�𝑦𝑦�
𝑄𝑡
𝑇𝑏
∗∗
∗∗∗ ∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗
∗∗∗
∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
𝑇𝑏
25
Hypothesized No. of CE(s) Eigenvalue Trace Statistic 0.05 Critical Value Prob.**None * 0.2838 165.4466 95.7537 0.0000At most 1 * 0.1686 94.6798 69.8189 0.0002At most 2 * 0.1250 55.5320 47.8561 0.0081At most 3 0.0612 27.2241 29.7971 0.0963At most 4 0.0336 13.8382 15.4947 0.0875At most 5 * 0.0306 6.5957 3.8415 0.0102 Trace test indicates 3 cointegrating eqn(s) at the 0.05 level
Hypothesized No. of CE(s) Eigenvalue
Max-Eigen Statistic 0.05 Critical Value Prob.**
None * 0.2838 70.7668 40.0776 0.0000At most 1 * 0.1686 39.1478 33.8769 0.0107At most 2 * 0.1250 28.3079 27.5843 0.0404At most 3 0.0612 13.3859 21.1316 0.4174At most 4 0.0336 7.2425 14.2646 0.4609At most 5 * 0.0306 6.5957 3.8415 0.0102 Max-eigenvalue test indicates 3 cointegrating eqn(s) at the 0.05 level * denotes rejection of the hypothesis at the 0.05 level**MacKinnon-Haug-Michelis (1999) p-valuesIncluded observations: 212 after adjustmentsTrend assumption: Linear deterministic trendSeries:
Unrestricted Cointegration Rank Test (Maximum Eigenvalue)
Table 3: Brazil Unrestricted Cointegration Rank Test (Trace)
𝑄𝑡 𝚤̃ 𝜋� 𝑚� 𝑇𝑏𝑦𝑦�
Hypothesized No. of CE(s) Eigenvalue Trace Statistic 0.05 Critical Value Prob.**None * 0.2699 158.9683 95.7537 0.0000At most 1 * 0.1633 91.9779 69.8189 0.0003At most 2 * 0.1141 54.0136 47.8561 0.0118At most 3 0.0582 28.1969 29.7971 0.0756At most 4 0.0472 15.4295 15.4947 0.0511At most 5 * 0.0238 5.1303 3.8415 0.0235 Trace test indicates 3 cointegrating eqn(s) at the 0.05 level
Hypothesized No. of CE(s) Eigenvalue
Max-Eigen Statistic 0.05 Critical Value Prob.**
None * 0.2699 66.9904 40.0776 0.0000At most 1 * 0.1633 37.9644 33.8769 0.0154At most 2 0.1141 25.8166 27.5843 0.0827At most 3 0.0582 12.7675 21.1316 0.4739At most 4 0.0472 10.2991 14.2646 0.1930At most 5 * 0.0238 5.1303 3.8415 0.0235 Max-eigenvalue test indicates 2 cointegrating eqn(s) at the 0.05 level * denotes rejection of the hypothesis at the 0.05 level**MacKinnon-Haug-Michelis (1999) p-valuesIncluded observations: 213 after adjustmentsTrend assumption: Linear deterministic trendSeries:
Tbale 4: South Africa Unrestricted Cointegration Rank Test (Trace)
Unrestricted Cointegration Rank Test (Maximum Eigenvalue)
𝑄𝑡 𝚤̃ 𝜋� 𝑚� 𝑇𝑏𝑦𝑦�
26
Hypothesized No. of CE(s) Eigenvalue Trace Statistic 0.05 Critical Value Prob.**None * 0.5890 350.2992 117.7082 0.0000At most 1 * 0.3177 161.8013 88.8038 0.0000At most 2 * 0.1936 80.7565 63.8761 0.0010At most 3 0.1095 35.1462 42.9153 0.2391At most 4 0.0374 10.5699 25.8721 0.8984At most 5 0.0116 2.4802 12.5180 0.9317 Trace test indicates 3 cointegrating eqn(s) at the 0.05 level
Hypothesized No. of CE(s) Eigenvalue
Max-Eigen Statistic 0.05 Critical Value Prob.**
None * 0.5890 188.4979 44.4972 0.0000At most 1 * 0.3177 81.0448 38.3310 0.0000At most 2 * 0.1936 45.6103 32.1183 0.0006At most 3 0.1095 24.5762 25.8232 0.0724At most 4 0.0374 8.0897 19.3870 0.8134At most 5 0.0116 2.4802 12.5180 0.9317 Max-eigenvalue test indicates 3 cointegrating eqn(s) at the 0.05 level * denotes rejection of the hypothesis at the 0.05 level**MacKinnon-Haug-Michelis (1999) p-valuesIncluded observations: 212 after adjustmentsTrend assumption: Linear deterministic trendSeries:
Table 5: Germany Unrestricted Cointegration Rank Test (Trace)
Unrestricted Cointegration Rank Test (Maximum Eigenvalue)
𝑄𝑡 𝚤̃ 𝜋� 𝑇𝑏𝑦𝑦�𝑚�
Hypothesized No. of CE(s) Eigenvalue Trace Statistic 0.05 Critical Value Prob.**None * 0.3676 200.4839 117.7082 0.0000At most 1 * 0.2625 103.3359 88.8038 0.0030At most 2 0.1041 38.7939 63.8761 0.8848
At most 3 0.0397 15.4778 42.9153 0.9976At most 4 0.0199 6.8942 25.8721 0.9945At most 5 0.0124 2.6414 12.5180 0.9162 Trace test indicates 2 cointegrating eqn(s) at the 0.05 level
Hypothesized No. of CE(s) Eigenvalue
Max-Eigen Statistic 0.05 Critical Value Prob.**
None * 0.3676 97.1480 44.4972 0.0000At most 1 * 0.2625 64.5420 38.3310 0.0000At most 2 0.1041 23.3161 32.1183 0.3955At most 3 0.0397 8.5836 25.8232 0.9900At most 4 0.0199 4.2528 19.3870 0.9957At most 5 0.0124 2.6414 12.5180 0.9162 Max-eigenvalue test indicates 2 cointegrating eqn(s) at the 0.05 level * denotes rejection of the hypothesis at the 0.05 level**MacKinnon-Haug-Michelis (1999) p-valuesIncluded observations: 212 after adjustmentsTrend assumption: Linear deterministic trendSeries:
Tble 6: United Kingdom Unrestricted Cointegration Rank Test (Trace)
Unrestricted Cointegration Rank Test (Maximum Eigenvalue)
All 4.6006 10 0.9162 All 10.509 10 0.3970Note:(1) *,**,*** denote rejection of the null hypothesis at the 10, 5, and 1% level of singificance, respectively.