ECONOMETRICS PROJECT REPORT topic factors affecting GDP of vietnam from 1995 to 2019

HANOI UNIVERSITY FACULTY OF MANAGEMENT AND TOURISM -o0o - ECONOMETRICS PROJECT REPORT Topic: Factors affecting GDP of Vietnam from 1995 to 2019 Tutor’s name: Ms Tran Thi Hoang Anh Tutorial class: Tutorial Group Phan Thi Hoang Yen | 1804040122 | Contribution: 100% Nguyen Minh Anh | 1804040010 | Contribution: 100% Nguyen Thuy Linh | 1804040067 | Contribution: 100% Do Minh Trang | 1804040110 | Contribution: 100% Hanoi, May 2021 LINK OF THE GROUP PRESENTATION VIDEO: TABLE OF CONTENTS I NATURE AND BACKGROUND OF THE STUDY Introduction Statement of the problem .6 Background of the statement Rationale for the study Research questions .7 II REVIEW OF LITERATURE III METHODOLOGY Definition of population .9 Sampling method use How the data was collected Research design used 10 Statistical tests 10 IV DATA ANALYSIS AND RESULTS 10 Descriptive Statistics 10 Interpretations: 13 2.1 Interpret coefficient .14 2.2. Coefficient of determination .14 Hypothesis Testing: 14 3.1 Testing the overall significance of all coefficient .14 3.2 Testing the individual partial coefficients 15 V CHECKING ERRORS IN THE MODEL .17 Multicollinearity 17 i 1.1 The nature 17 1.2 Consequences 18 1.3 Detection .18 1.4 Remedial measures 19 Heteroscedasticity 19 2.1 The nature 20 2.2 Consequences 20 2.3 Detection .20 2.4 Remedial 22 Autocorrelation 23 3.1 The nature 23 3.2 Consequences 23 3.3 Detection .23 3.4 Remedial 25 Summary of checking errors of the model 26 VI SUMMARY, CONCLUSION AND RECOMMENDATIONS 27 Summary and conclusion: 27 Recommendation 27 APPENDIX 28 REFERENCES LIST 29 ii TABLE OF FIGURE Figure 1: Statistics Description Figure 2 : Correlation matrix .9 Figure 3: Covariance matrix Figure 4: Relationship between GDP and its factors 10 Figure 5: Estimation of the best model 11 Figure 6: Variance Inflation Factor 16 Figure 7: White’s Heteroscedasticity test 18 Figure 8: White Heteroskedasticity-consistent standard errors & covariance test 22 Figure 9: Breusch-Godfrey serial correlation LM test for AR (2) .25 Figure 10: Newey-West HAC standard errors & covariance 26 iii ACKNOWLEDGEMENTS The research would not be done completely without any assistance Thus, we appreciate every support and motivation during the time doing this research Thanks to the knowledge that we learned from the Econometrics textbook we absorbed the basics which would help us build up the idea of this research Moreover, we want to give acknowledgment to the author of the prior research which gave us detailed information in our study Finally, we deeply thank two dedicated teachers, Ms Dao Binh, Mr Pham Hung and Ms Hoang Anh iv I NATURE AND BACKGROUND OF THE STUDY Introduction The global economy is undergoing significant changes, the countries around the world are tending to integrate globally, helping to grow the economy through the exchange and trading of goods not only in the region but also in other continents Economic development affects the prosperity as well as political factors of nations, Vietnam is no exception After more than 20 years of revolution, first of all, is "economic thinking", shifting from a centrally planned economy to a socialist-oriented market economy, promoting industrialization, modernizing, proactively and actively integrating into the world, Vietnam has made many remarkable achievements in the domestic economy From an agricultural, backward, small-scale economy, with a GDP of only 14 billion USD and GDP per capita of only about 250 USD in the early years of Doi Moi, Vietnam has come out of poverty, moved to implement and step up industrialization and modernization of the country By 2019, Vietnam had official relations with 189/193 countries of the United Nations (compared with 11 countries in 1954); having economic, trade, and investment relations with over 224 countries and territories worldwide; has 16 strategic partners, 11 comprehensive strategic partners; join more than 500 bilateral and multilateral agreements in many fields (including 16 FTAs); 71 countries have recognized Vietnam as a market economy, etc Vietnam has opened its doors, became a member of ASEAN, APEC, the WTO, and many other international organizations participating in many free trade institutions, proactively and actively contribute to building and shaping multilateral institutions, becoming a reliable partner and a responsible member of the international community, integrating more deeply and broadly into the global economy, gradually expanding integration into all fields of politics, defense, security, and culture – society, etc This is a very important step and opens up a promising economy Economic growth takes place, it is reflected in the increasing and stable GDP growth rate for a long time, the economy will have many great achievements Thus, the more stable the income and living standard of the people, the more developed the country is Therefore, economic growth is considered an attractive issue in economic research, it is the focal point to reflect the changing face of the national economy Statement of the problem  We can easily see that economic development has a great influence on life, culture, and political activities in countries around the world in general and Vietnam in particular Economic development when looking at a country's GDP, the components of GDP generally are the main factors in the economy To evaluate a country's economy, economists evaluate the gross domestic product GDP Background of the statement First of all, GDP is an economic term that stands for the English phrase Gross Domestic Product This term means the gross domestic product (also known as gross domestic product) This is an index given to assess the overall and generalized growth rate of each country's economy and assess the development level of that country GDP is the value calculated according to the quantity of all goods and all services formed in a territory in a given period of time The time to calculate the value of GDP in each type of goods or service is usually from months, months, months or year depending on each specific sector GDP is the index to calculate the value of all products and services in the domestic economy, including foreign companies based in Vietnam Constructed in a three-step sequence, the GDP is calibrated to four sub-indices, reflecting economic factors that influence the development of the economy of Vietnam: • Population (P): The population is a collection of people living in a certain geographical area or space that is a valuable source of labor for socio-economic development, often measured by census and expression by population chart Population is both the production force and the consumer force The size of the population affects the workforce, which will give the country the ability to comprehensively develop economic sectors, while at the same time having profound labor expertise, creating conditions for productivity improvement labor, thereby promoting economic and social development • Investment (I): The total personal investment includes the business's expenditure on equipment and factories or the construction and purchase of a new home by the household In addition, inventories, when being put into stock, and not yet sold, are also included in GDP In macroeconomics, it is only necessary to increase capital to strengthen future production capacity Private gross domestic investment can have a major effect on accelerating economic growth by creating new businesses that attract more workers, thereby solving economic difficulties, society's unemployment More than women, private investment also facilitates increased budget revenue in the form of taxes • Exports (X): Domestically produced goods that are sold abroad (the proceeds from the sale of goods and services abroad - which increases GDP) • Imports (M): goods that are produced abroad, but purchased for domestic demand (the amount paid abroad by the purchase of goods and services - reduces GDP) When we export, it will reduce the net worth bringing to the economy and it will increase the net worth in the economy if we an import Rationale for the study GDP accurately reflects the economy of a country, it clearly shows the change and the level of equilibrium that keep the main factors that make up GDP and its index of change as well as the efforts of the government to it factors that make up GDP over the years It is essential to determine the factors influencing GDP and the correlation among these factors Building the model based on these factors, we can know our strengths and weaknesses in the economy; That has the potential to carry out more appropriate economic adjustment strategies, helping to increase GDP over the years and develop the economy Research questions The relative growth rate of GDP can be affected by a number of factors, some of which show an inverse relationship while other factors show a direct relationship.  This assignment is devoted to analyzing the extent to which those factors affect GDP growth in Vietnam The research space of the topic is within the entire economy of Vietnam, because of the limitation in data sources, the research period is in the period 1995 - 2019 In this study, we will present the procedures in collecting data and the process of making our conclusion about: - What is the relationship between GDP and four sub-indices? How these determinants affect the economy? - Are there any connections among these determinants? - Are there any errors shown while running the model? II REVIEW OF LITERATURE Prior research related literature Before taking the projects, we looked for other research to see how the GDP was studied through previous research: No Author/Year Research Result Limit Karen Dynan GDP as a Macroeconomics Practical Measurement No (2018) of Theory Methods Subject Measure of theory: GDP Theoretica Economic l Well-being  Random variables GDP specific through years evaluate well-being  the analysis to of the factors affecting GDP Alex Reuben The Macroeconomics Cross Kira (2013) theory: GDP - tabulation influence of factors Consumption on UK’s and Export - Analyzing factors No specific affecting GDP analysis in Developing of GDP from Countries: The factors 1965 to case 2010 Dhiraj Tanzania Jain The Macroeconomics Cross K Sanal Nair influence and Vaishali of Jain (2015) on Net FII UK’s equity, Net FII GDP from debt, Import and 1965 to Export the impact of specific various analysis macroeconomic of factors on GDP factors components 2010 GDP - To investigate No theory: GDP - tabulation  factors FDI, of affecting affecting GDP It is clearly seen that GDP in Vietnam is a crucial index for both domestic and foreign economists The economists and organizations desire to know which element is the most important factor affecting the Vietnamese economy Overall, the three reports have limitations in that they not contain specific indicators affecting the GDP and taking the examination within 25 years III METHODOLOGY Definition of population The scope of the research is the GDP which includes population, investment, exports and imports Sampling method use To be more precise, we evaluated the GDP by collecting the statistics of 25 years in Vietnam Thus, the sample size is 25 Therefore, to demonstrate the data and organize them properly, we used Microsoft Excel Ho: All variables have no effect on GDP (β2 = β3 = β4 = β5 = 0) H1: At least one variable has effect on GDP (β2≠0, β3≠0, β4≠0 or/and β5≠0) F-statistic In the Eviews table above, we obtain: Test statistic: F-stat = 1392.059 Critical value: F cα ,k−1 ,n−k = F c0.05,4,6 = 4.53 Decision rule: If   F-stat > F c => Reject Ho Compare:  1392.059 > 4.53 Decision: Reject Ho 3.2 Testing the individual partial coefficients Through the previous part, we were concerned with the significance of all estimators And now we use a t-test to test a hypothesis about any partial regression coefficient These tests aim to check whether each independent variable is significant or not Intercept Hypothesis coefficient : H0: β1 = H1: β1 ≠ Critical value: c tα ,n−k There is not enough evidence to = t c0.025,6 = 2.447 statistically significant with 95% of confidence level Test statistic: t* = conclude that the intercept coefficient ^ β 1−0 −3053843 = 1537421 Se ( ^ β 1) = -1.9863 Conclusion: | t* | < t c (1.9863 < 2.447) ⟶ Do not reject Ho 15 Population Hypothesis : H0: β2 = H1: β2 ≠ Critical value: c tα ,n−k There is enough statistical evidence to = t c0.025,6 = 2.447 conclude that Population has an effect on GDP with 95% of confidence level Test statistic: t* = ^ β 2−0 0.040863 = ^ Se ( β 2) 0.020142 = 2.0286 Conclusion: | t* | > t c (2.0286 > 2.447) ⟶ Reject Ho Investmen Hypothesis t : H0: β3 = H1: β3 ≠ Critical value: c tα ,n−k There = t c0.025,6 = 2.447 not enough statistical evidence to conclude that Investment has an effect on GDP with 95% of confidence level Test statistic: t* = is ^ β 3−0 1.473427 = ^ Se ( β 3) 0.882330 = 1.6699 Conclusion: | t* | < t c (1.6699 < 2.447) ⟶ Do not reject Ho Export Hypothesis Critical value: There is enough statistical evidence to : conclude that Export variable has an effect on GDP with 95% of confidence 16 t cα β4 = β4 ≠ ,n−k = t c0.025,6 = 2.447 level Test statistic: t* = ^ β 4−0 0.763855 = ^ Se ( β 4) 0.183876 = 4.1542 Conclusion: | t* | > t c (4.1542 > 2.447) ⟶ Reject Ho Import Hypothesis : Critical value: c tα β5 = β5 ≠ ,n−k There = t c0.025,6 = 2.447 not enough variable has an effect on GDP with ^ β 5−0 −0.401301 = 0.327197 Se ( ^ β 5) = -1.2265 Conclusion: | t* | < t c (1.2265 < 2.447) ⟶ Do not reject Ho V CHECKING ERRORS IN THE MODEL Multicollinearity One of ten assumptions of the classical linear regression model (CLRM) is that there is no multicollinearity among regressors (Assumption 10) This might be because the existence of 17 statistical evidence to conclude that Import 95% of confidence level Test statistic: t* = is multicollinearity leads to less accuracy results of the regression coefficients and reduces the reliance of the model.   1.1 The nature   Multicollinearity exists when there are perfect linear relationships among independent variables of the regression model This exact relationship happens if the following condition is satisfied:   λ1X1+ λ 2X2+ …+ λkXk vi = 0   vi is a stochastic error term  λ1, λ2, …, λk constant Test for multicollinearity must be done to identify whether there are some functional relationships among explanatory variables so that improve the precision and accuracy for the model.   1.2 Consequences There are several consequences when imperfect multicollinearity exists, namely:  Large variances and covariances which make the estimation less accurate  The estimation confidence intervals tend to be much wider, increasing the chance to  accept “zero null hypothesis”  The t-statistics of coefficients tend to be statistically insignificant  The R2 can be very high  The OLS estimators and their standard errors can be sensitive to small changes in  the data.   1.3 Detection In order to find out how much the variance (the square of the estimate's standard deviation) of an estimated regression coefficient is increased because of collinearity, we use a method 18 which uses the variance inflation factor: VIF = 1−Ri Figure Variance Inflation Factor As it can be seen from the table, all the VIF are greater than 10, our variables are high collinear Therefore, there is a multicollinear relationship between regressors 1.4 Remedial measures To sum up, our model incurs a significant issue related to multicollinearity (except BIAS) However, according to Blanchard (1697), “Multicollinearity is essentially a data deficiency problem and sometimes we have no choice over the data we have available for empirical analysis” Therefore, doing nothing occasionally is the best we can given the set of data.   19 Heteroscedasticity Heteroskedasticity (unequal conditional variance of error terms), is the most popular error as constructing models with cross sectional data This contrasts Assumption about homoscedasticity Compared to multicollinearity, heteroskedasticity contributes to a more serious problem 2.1 The nature Heteroscedasticity exists if variances of error terms in any model are not constant according to changes in explanatory and explained variables Symbolically, Var(ui ) = E(ui2) = σ i2 is not constant (for i = 1, 2, , n) This indicates that the disturbance for each of the n-units is drawn from a probability distribution that has a different variance There are in fact both formal and informal methods to test for the existence of heteroscedasticity 2.2 Consequences  OLS estimators are still linear and unbiased  Var( ^ β 1)s are not minimum  Var( ^ β 2) =  The estimated variances and covariances are biased and inconsistent  t and F statistics are unreliable σ i2 σ2 ^ instead of Var( ) = β ∑ x2 ∑ x2 2.3 Detection We use White’s heteroscedasticity test (without cross term, with cross term case would eat up a lot of degree of freedom) to find out whether there exists heteroscedasticity in our OLD model or not) 20 Figure 7: White’s Heteroscedasticity test Hypothesis: H 0: Homoscedasticity Var (ui ) = σ H 1: Heteroscedasticity Var (ui ) = σ Test statistic: W = n × R2= 25 × 0.887065 = 22.176625 Critical value: X 2α , df = X 20.05,14= 23.685 Decision rule: If W ¿ X α, df => Reject H Compare: X 20.05,14= 23.685 > W = 22.176625 => Do not Reject H 21 Conclusion: There is not enough evidence to infer that heteroscedasticity exists from this model at 5% level of significance 2.4 Remedial With hypotheses testing above, our model does not violate the Assumption of CLRM but after reducing the effect of Heteroscedasticity by transforming the model to a log-log model and using “White Heteroskedasticity-consistent standard errors & covariance” method which is a robust method This action helps us have a better model with the result from Eviews below: Figure 8: White Heteroskedasticity-consistent standard errors & covariance test In this new model, we have the W-statistic (n R2) = 6.339463 < X 20.05,4= 9.49 which shows that the model still be a Homoskedasticity 22 Autocorrelation The purpose of autocorrelation test is to check whether the linear relationship between errors exists or not, which breaks the Assumption of CLM (cov( ui , u j | x i, x j ) = 0) Certainly, the importance of the autocorrelation can be considered the same as heteroskedasticity test The primary reason is that the standard error of coefficient becomes larger and not the minimum value of variances when autocorrelation occurs 3.1 The nature When assumption cov(ui , u j) = 0, i≠j is violated, then comes the sao-called autocorrelation which can be performed by cov(ui , u j) ≠ 0, i≠j, which means there is serial correlation among the disturbances entering into the population regression function.  3.2 Consequences  The estimated coefficients remain unbiased   Var( ^ β 1) is no longer the smallest Therefore, its standard error also becomes large  The usual t and F tests of significance are no longer valid  ∑ ui 2 ^ The residual variance σ = is likely to underestimate the true σ n−2  R-squared is more likely to be overestimated 3.3 Detection a Durbin-Watson Test AR(1): Step 1: Ho: There is no positive autocorrelation existing            Ha: There is positive autocorrelation existing Step 2: Test statistic            Durbin-Watson stat: d* = 0.561738 (from Eviews) Step 3: Critical value: n = 25, k’= 4 => d L=0.832, d U = 1.521 Step 4: Decision rule           Reject Ho if < DW stat < d L 23           Do not reject Ho if - d L< DW stat < ord U < DW stat < - d U           Inconclusive if d L< DW stat < d U or - d U < DW stat < - d L Step 5: Conclusion: Since < DW stat (0.561738) X 20.05 ,df           Do not reject Ho if LM < X 20.05 ,df Step 4: Value of test statistic:            The result of the BG test for the AR (2) is:            LM = (n - 2) × R2 = 12.20301            X 20.05,2 = 5.991 Step 5: Conclusion: LM > X 20.05 ,df (5.991), we reject Ho There is enough evidence to conclude that there is high-order autocorrelation existing in order.  24 Figure 9: Breusch-Godfrey serial correlation LM test for AR (2) 3.4 Remedial After testing autocorrelation, we defined that there was high-order autocorrelation existing in our model, then we decided to adjust the standard errors of regression coefficients by using the Newey-West method In this case, our Newey-West method handles autocorrelation with lags up to 2, and it is assumed that lags larger than can be ignored.  Here we have the result from Eviews: 25 Figure 10: Newey-West HAC standard errors & covariance As we can see in figure 10, the OLS standard errors are not significantly different from figure above This suggests that although there is evidence of correlation based on some tests, the problem is that the autocorrelation does not seem too severe This probably due to the fact that the correlation detected in the term of interference, between 0.32 and 0.35, may not be too high Summary of checking errors of the model After checking three errors that may have in our model we can conclude that there are multicollinearity and autocorrelation error To detect the multicollinearity, we have to use methods: the first method compares the R-square between auxiliary and original regression error haven’t detected but with the VIF method the problem has come out However, as we discuss with three reasons above we decide not to fix the multicollinearity To detect autocorrelation, we use Newey-West to adjust the standards errors of regression coefficients and the result is not too severe then we decide to keep the old model 26 VI SUMMARY, CONCLUSION AND RECOMMENDATIONS Summary and conclusion:  GDP can be considered as a criterion to evaluate a country's economy, it allows investors, large companies, corporations to seek opportunities and plan to invest, participate in the economic activity of that country In order to understand in details the factors affecting the change of GDP in general and GDP in Vietnam in particular from 1995 to 2019, we have conducted this research paper, including regression function and performed essential tests Based on our model, there are variables influencing Vietnam’s GDP: Population, Investment, Imports, Exports After using t-test to check the significant off each variable, we discover that Investment (I) and Import (M) are insignificant with only a 5% level of significance In terms of the multicollinearity test, we used variance factor VIF and received the result that all VIF’s are higher than 10 Then, our group decided to take the heteroscedasticity test and concluded that there is not enough proof to confirm that heteroscedasticity exists The result of Durbin-Watson Test and Breusch-Godfrey Test for autocorrelation signifies that there is existence of autocorrelation in two orders Therefore, we used Newey-West and the result is not too severe for autocorrelation Recommendation   Below are some recommendations to rise the GDP of Vietnam: First, we need to be aware of the importance of GDP as well as how its values affect the economy of Vietnam in general and countries in the world that speak Moreover, we need to statistically and clearly analyze data of factors affecting the change in GDP over the years (can be monthly or quarterly if necessary) We aim to increase GDP in Vietnam over the years, so we need to pay more attention to investment Our country's economy needs to expand and improve policies to create favorable conditions for foreign corporations and companies to invest in the country's economy 27 On the other hand, we need to care about import and export Increasing domestic goods circulation, actively exporting commodities to foreign markets, and reducing imports are factors that help GDP grow over the years APPENDIX Appendix: Data source of GDP - Dependent variable (Y) GDP OF VIETNAM FROM 1995-2019 (billion VND) Year GDP (Y) Population (K) Investment (I) Export (X) 1995 227,473.92 74,910,460 72,447.00 1996 273,692.70 76,068,730 87,394.00 1997 299,981.70 77,133,210 108,370.00 1998 353,321.85 78,115,710 117,134.00 1999 401,690.74 79,035,870 131,171.00 2000 433,335.87 79,910,410 151,183.00 2001 484,849.29 80,742,500 170,496.00 2002 531,955.94 81,534,410 200,145.00 2003 596,815.22 82,301,660 239,246.00 2004 716,399.56 83,062,820 290,927.00 2005 916,018.90 83,832,660 343,135.00 2006 1,025,049.17 84,617,540 404,712.00 2007 1,225,231.38 85,419,590 532,093.00 2008 1,646,846.69 86,243,410 616,735.00 2009 1,809,149.00 87,090,000 708,826.00 2010 2,157,828.00 87,970,000 830,278.00 2011 2,779,880.00 88,870,000 924,495.00 2012 3,245,419.00 89,800,000 1,010,114.00 2013 3,584,262.00 90,750,000 1,094,542.00 2014 3,937,856.00 91,710,000 1,220,704.00 2015 4,192,862.00 92,680,000 1,366,478.00 2016 4,502,733.00 93,640,000 1,487,638.00 2017 5,005,975.00 94,600,000 1,670,196.00 2018 5,542,331.90 95,540,000 1,857,061.00 2019 6,037,348.00 96,460,000 2,046,838.00 Source: General Statistic Office Vietnam 28 59,775.53 80,541.60 97,848.30 121,072.14 161,326.08 200,855.55 222,910.52 253,446.73 304,435.66 417,968.08 515,633.15 615,072.74 768,574.95 1,041,385.91 1,313,214.90 1,661,444 2,228,831 2,634,172 3,036,757 3,454,993.30 3,726,384.10 4,061,358.40 4,947,727.80 5,605,037.90 6,117,296.30 Import (M) 89,460.35 123,687.30 124,612.43 149,249.59 217,342.14 239,747.11 299,369.34 380,650.20 503,914.58 587,728.33 693,296.60 992,068.01 1,340,901.68 1,608,822.40 1,608,822.40 1,951,287.80 2,455,245 2,616,949 3,036,750 3,400,529.30 3,812,845.70 4,024,503.20 4,903,951.90 5,455,186.00 5,856,323.94 REFERENCES LIST General Statistics Office Retrieved from: https://www.gso.gov.vn/default_en.aspx?tabid=491 Gross Domestic Product (GDP) Retrieved from: https://www.investopedia.com/terms/g/gdp.asp “Cải cách kinh tế giúp làm thay đổi diện mạo kinh tế Việt Nam” Retrieved from: http://hdll.vn/vi/thong-tin-ly-luan/cai-cach-kinh-te-giup-lam-thay-doi-dien-mao-kinh-te-vietnam.html “Kinh tế Việt Nam liệu có “hay ho” số liệu tính lại?” Retrieved from: https://www.bbc.com/vietnamese/vietnam-50999289 Components of GDP Explained Retrieved from: https://www.thebalance.com/componentsofgdpexplanationformulaandchart3306015#:~:text=The %20four%20components%20of%20gross,being%20spent%20in%20that%20economy 29