RESEARCH TOPIC UNIVERSITY OF ECONOMICS HOCHIMINH CITY VIETNAM INSTITUTE OF SOCIAL STUDIES THE HAGUE THE NETHERLANDS VIETNAM – NETHERLANDS PROGRAMME FOR M A IN DEVELOPMENT ECONOMICS FACTORS AFFECTING T[.]
UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL HOCHIMINH CITY STUDIES VIETNAM THE HAGUE THE NETHERLANDS VIETNAM – NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS FACTORS AFFECTING THE WORLD’S GOLD PRICE: AN ARDL APPROACH BY VU THUY DUONG MASTER OF ARTS IN DEVELOPMENT ECONOMICS HOCHIMINH CITY, OCTOBER 2013 UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES HO CHI MINH CITY THE HAGUE VIETNAM THE NETHERLANDS VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS FACTORS AFFECTING THE WORLD’S GOLD PRICE: AN ARDL APPROACH A thesis submitted in a partial fulfillment of the requirements for the degree of MASTER OF ARTS IN DEVELOPMENT ECONOMICS By VU THUY DUONG Academic supervisor: Dr CAO HAO THI HOCHIMINH CITY, OCTOBER 2013 Acknowledgement The thesis could not be completed without considerable supports from my academic supervisor, Dr Cao Hao Thi, who provided me valuable instructions and comments throughout the process of this thesis From bottom of my heart, I would like to give my sincerest thanks to him I am also grateful to Dr Le Van Chon and Dr Truong Dang Thuy for their enthusiasm help about Econometrics techniques and useful advice Therefore, I can overcome many obstacles to complete my research I also take this opportunity to express my thanks for my colleagues They provided me many assistance and encouragement during the time I the research Last but not the least, I would like to thank my family members for their love, encouragement, support for me to finish the Master course as well as the thesis Abstract The paper focuses on investigating factors affecting global gold prices in the shortrun and long-run with daily data from January 2007 to December 2012 By applying autoregressive distributed lag (ARDL) bound test, the empirical results show that there is no evidence of long-run relationship among London gold price, West Texas Intermediate (WTI) crude oil sport price, US dollar index and S&P 500 However, when financial crisis is taken into the research as a dummy variable, the results reveal that financial crisis affects the relationship of gold price with oil price, US dollar index and S&P 500 and cannot conclude that long-run relationship among them existed Therefore, the State Bank of Viet Nam cannot base on the movement of those variables to forecast the movement of world gold price for making their decision in selling or buying gold Table of Contents CHAPTER 1: INTRODUCTION .6 1.1 Problem Statements 1.2 Research objectives 1.3 Research questions 1.4 Scope of the research 1.5 Structure of the thesis .9 CHAPTER 2: OVERVIEW OF VIETNAM’S GOLD MARKET 11 2.1 The national gold brand of Vietnam .11 2.2 The connection of Vietnam’s gold price to global gold price 11 2.2.1 Domestic gold’s Cost price .11 2.2.2 Domestic gold’s market price 12 2.2.3 The connection between domestic and global gold markets 12 2.2.4 The big gap still exists between the two gold markets .13 CHAPTER 3: LITERATURE REVIEW 15 3.1 The relationship between gold and oil prices 15 3.2 The relationship between gold price and US Dollar exchange rate .16 3.3 The relationship between gold price and stock market 17 CHAPTER 4: RESEARCH METHODOLOGY 21 4.1 Research process 21 4.2 Model establishment .21 4.3 Data collection 21 4.4 Data analysis 24 4.5 Analysis method .25 4.5.1 Stationary and unit root test .25 4.5.2 Cointegration test .26 CHAPTER 5: RESEARCH RESULTS 31 5.1 Descriptive statistics .31 5.2 Correlation matrix 32 5.3 Stationary and unit root test 33 5.4 Cointegration analysis 34 5.4.1 Optimal lag length 34 5.4.2 Serial correlation test .35 5.4.3 Dynamic stability test 35 5.4.4 Bound tests .36 CHAPTER 6: CONCLUSION AND POLICY IMPLICATIONS 38 6.1 Main findings .38 6.2 Policy implications 39 6.3 Limitation .40 6.4 Future research .40 References 41 Appendix A: Lag structure choosing .46 Appendix B: Bound test result 50 List of Figures Figure 4.1: Research process 22 Figure 4.2: Analysis method .25 Figure 5.1: Inverse Roots 36 List of Tables Table 2.1: Correlation matrix between SJC and London’s gold price 13 Table 2.2: Correlation testing between SJC and London’s gold price .13 Table 3.1: Summary of empirical studies in Literature review .18 Table 4.1: Asymptotic critical value bounds for the F-statistics 29 Table 5.1: Descriptive statistics of all series .31 Table 5.2: Correlation matrix 32 Table 5.3: Unit root test for stationary at level 33 Table 5.4: Unit root test for stationary at first difference 33 Table 5.5: VAR lag order selection criteria 34 Table 5.6: Serial correlation test’s result 35 Table 5.7: Bounds test procedure results without crisis interaction 36 Table 5.8: Bounds test procedure results with crisis interaction 36 Abbreviations ADF: Augmented Dickey-Fuller AIC: Akaike Information Criterion ARDL: Autoregressive Distributed Lag CPI: ECM: Consumer Price Index Error correction models EIA: U.S Energy Information Administration PP: Phillips-Perron SBV: SJC UECM: US: VAR: VND: State Bank of Viet Nam Saigon Jewelry Company Limited Unrestricted error correction model United States Vector Autoregression Estimates Vietnam Dong CHAPTER 1: INTRODUCTION 1.1 Problem Statements There is much attention recently on gold, partly due to the surges in its price and the increase in its economic uses Gold has a critical position among the major precious metals Gold is not only an industrial commodity but also an investment asset which is commonly known as a “safe haven” (Baur & Lucey, 2010; Coudert & Raymond-Feingold, 2011) to avoid the increasing risk in the financial markets However, gold prices have been remained unabated, even accelerated further in recent years Since the beginning of financial crisis in August, 2007 to December, 2012, the nominal gold price has increased 146.65% For central banks, gold keeps an important position in their reserve asset Its role is increasingly enhanced from 2009 when there was so much worry about the health of US economy Central banks increase to buy gold due to they want to reduce their reliance on the US dollars as a reserve asset, and to encourage borrowings and to ensure interest rates on these loans were reduced Bellowing is some evidence recently relating to the trend of buying gold from central banks: i “Asked what the most important reserve asset would be in 25 years, about half of officials polled by UBS said the US dollar but 22 percent pointed to gold” (Farchy & Blas, 2010) This is showed that the central banks’ demand for gold will remain strong over time ii And the report of the World Gold Council published on August 14, 2012, mentioned that: “The second quarter was another period of significant purchasing by official sector institutions, with demand accounting to 157.5 tonnes This was a record quarter for central bank buying since the sector began recording net purchases in Q2 2009 and was more than double the 66.2 tonnes of purchases made in the same period of 2011 Purchases in the first half of the year totaled 254.2 tonnes, 25% up on 203.2 tonnes from the same period last year The official sector accounted for 16% of overall Q2 gold demand”(“Gold Demand Trends Q2 2012”,2012) It is clearly showed that the role of gold for central banks is even expanding In addition, there are some notable policies in relation to gold that central banks and nations are applying to ensure that gold accumulated either in their citizens’ hands or in central bank hands (Phillips, 2012): i China putted a ban on exporting gold to ensure that all gold that enters the country stays in the country; ii If the country can produce gold, its central bank will be the directly purchaser For example: Russia & Kazakhstan bought its locally produced gold in March, 2012; iii There is the fact that the trend of buying gold mostly comes from central banks of emerging and newly wealthy nations Their gold holdings are very small in comparison to central banks of US and Europe In Viet Nam, the domestic gold prices got so much fluctuation since September 2009 Especially in 2012, the gap between global and domestic prices had been largely widen, sometimes reaching VND million per tael (a tael is equal to 1.2 ounces) when the demand for gold is larger than the supply for gold due to the “big guy banks” This has caused serious problems to the exchange rate as well as economy Khanh (2010), the General Director of Sai Gon Gold and Silver ACBSJC Joint Stock Company, argued that the increasing in gold price will have both direct and indirect impact to the USD/VND exchange rate, indirect impact to the CPI, affect monetary policy, stock market and real estate market, etc To solve the problem of gap or to stabilize the gold market that is also the requirement of the National Assembly of the Socialist Republic of Vietnam from Resolution No 51/2010/QH12 on the 2011 Socio-Economic Development Plan, there are some notable legal documents issued by the Government and The State Bank of Vietnam (SBV) from 2010 such as: (i) Circular No 01/2010/TT-NHNN dated 06/01/2010 about closing the gold trading floor and terminating all activities of gold trading on foreign account; (ii) Circular No 22/2010/TT-NHNN dated 29/10/2010 about not allowing the commercial bank to convert gold into paper currency and lend for gold trading activities; (iii) Decree No 24/2012/ND-CP dated 03/04/2012 about using gold bars as a tool of payment will be illegal and the State will keep a monopoly on gold bar production, export and import of raw materials for gold bar production via the SBV; (iv) Circular No 16/2012/TT-NHNN dated 25/05/2012 guiding some articles of Decree No 24/2012/ND-CP Yet, those documents have limited the gold supply to the market while the gold demand is still large that make gold prices still “crazy” when global gold price increases In addition, according to the Decision No 16/2013/QD-TTg issued by the Prime Minister on 04/03/2013, gold bullion would be bought and sold by the SBV as needed to keep the gold prices stable and SBV, based on its monetary policies, would buy additional gold bullion from other countries for the aim of controlling foreign reserves This decision is highly appreciated due to it will be an opportunity for the SBV to fully perform the role of “conductor” to stabilize gold prices So, it is necessary that SBV should choose appropriate time to buy gold from abroad in order to increase reserve asset or to boost the supply side In other words, SBV should find out tools to forecast the movement of global and local gold prices when other factors changed From the above features, the paper will focus on analyzing factors (that are addressed in the literature) affecting global gold prices 1.2 Research objectives The main objectives of this paper include: i To investigate the relationship between local and global gold prices; ii To identify variables that affects global price of gold in the short-run and longrun periods; 60 tons per year The difference between the two markets was very low, but the domestic gold market under instability, gold fevers occurred frequently, people rushed to buy or sell gold, speculative activities seriously affected foreign exchange markets, price indices and macroeconomic stability Therefore, the government had banned and officially terminated the operation of gold floors, gold trading activities in foreign accounts 2009-2012 stage: the difference between the two markets remained low, but higher than the above stage The domestic market was still under instability that causing negative impact to exchange rates, macro-economy, etc but at the lower level 2012-2013 stage: New legal framework has been developed and come into effected The State has got a monopoly on gold import and gold bar production Gold smuggling is also controlled tightened The features of this period are: the gap between two market are much higher than the two periods, but the domestic market is more stable, speculation pushed back, no scenes of people of people rushing to buy gold, “goldenization” in the national economy is restrained, stable macroeconomy, etc In conclusion, the domestic gold price correlated with the international gold price in the researching period The difference between two markets still exists due to domestic gold demand and supply as well as State’ policies in each periods 14 CHAPTER 3: LITERATURE REVIEW The aim of this chapter is to give an overview of literature about the research problem in a logical manner to make the thesis going into the right direction Three sections included in the Chapter The first section gives literatures about the relationship between gold and oil prices The second section includes literatures on the relationship between gold price and US Dollar exchange rate The third section will present studies on the relationship between gold price and stock market The method adopted to use in this thesis will be mentioned in the conclusion of the Chapter 3.1 The relationship between gold and oil prices There are considerable studies on the relationship between gold and oil prices However, results on the relationship are still mixed Some studies explained the relationship through two following channels: i Firstly, oil price affects gold price through the inflation-channel (Narayan et al., 2010; Malliaris et al., 2011) When oil prices increase, the general price level will rise (Cunado et al., 2003, 2005) It follows that when the general price level (or inflation) rises, the price of gold will go up because gold is also a kind of goods This leads to the possibility that gold could be used to hedge against inflation (Jaffe, 1989); hedge against the dollar (Capie et al., 2005) ii Secondly, oil price affects gold price through the export revenue channel (Melvin & Sultan, 1990) With the assumption that gold accounts for a significant share in asset of the international reserve portfolio of several countries, including the oil producing countries The rise in oil prices (and hence oil revenues) may have implications for the rise of gold prices This can be explained that the increase of revenues from oil enhances the demand for gold and therefore, the price of gold will be increased Empirically, Narayan et al (2010), via inflation channel, proved that the relationship between gold and oil prices exists in both short-run and long-run For 15 the short-run, they used an ordinary least squares approach to test the relationship between gold and oil prices via inflation over the period 1963-2008 for the United States and found that an increase in oil prices led to inflation, an increase in inflation led to a rise in the gold price, and a rise in oil price led to a rise in the gold price For the long-run, they applied a structural break cointegration test proposed by Gregory and Hansen (1996) which has more advantage than the traditional cointegration test proposed by Engle and Granger (1987) because it allows for a regime shift The data they used is for the period 1995 to 2009 They found that gold and oil spot and future markets up to the maturity of 10 months were cointegrated; oil prices can be used to predict gold prices and vice versa However, Harmmoudeh et al (2008) and Sari et al (2010), by their empirical studies, not have the same results as Narayan et al (2010) They found that oil price does not force gold price in the long-run and vice versa When prices of gold fluctuates due to affected by factors related to the jewelry industry, and subject to intervention by central banks as part of their foreign reserve policies to influence exchange rates, it seems not to have anything related to oil price In addition, Malliaris et al (2011) did not found any evidence of cointegration between gold and oil prices when taking Johansen test as well as Granger causality test for the period from January 4, 2000 to December 31, 2007 But, gold price can be used to predict oil price in the short-run very well and vice versa when they took neural network methodology 3.2 The relationship between gold price and US Dollar exchange rate The price of gold and the value of the US dollar have been argued that they tend to move in opposite directions In particular, when gold went up, the dollar went down, and vice versa The reason is that a weaker dollar can make worries about inflation That will encourage investors to turn to gold to hedge against inflation; and when the dollar strengthens, gold price will usually fall Relating to the relationship, Hammoudeh et al (2008) found that an increase in the price of gold has implication for a depreciation of the US dollar versus the major 16 currencies in both short-run and long-run, not vice versa In contrast, Kim and Dilt (2011) found that the value of dollar and the gold price has negative relationship But value of dollar has Granger-causality to gold price, not vice versa 3.3 The relationship between gold price and stock market It is also logically to expect that the inverse relationship exists between gold price and stock price When stock price rises, investors will get more money from stock market and then they will want to sell their gold to invest more in the stock market This will cause the price of gold decreased Smith (2001) examined the short-term and long-term relationships between four gold price series and six different US stock price indices over the 1991-2001 period He found that there is a negative Granger causality from US stock index returns to gold returns in short-run, not vice versa However, Gilmore et al (2009) did not have the same result They only found that the unidirectional causal effect exists only in short-run In addition, Wang et al (2010), by applying Granger causality analysis, found that gold prices and Taiwan’s stock prices are independent It means that they cannot affect each other And Bhunia & Das (2012) also have other view By applying Granger causality analysis, they found that stock market can be used to predict gold prices in India and vice versa In summary, a general judgment is that researches’ results are still mixed That is, while some papers found causality existing between variables, others figure out no causality and/or bi-directional causality between variables The differences among them may be lie on the sample periods, research methodologies, and variables used Therefore, the paper will continue to test against the relationship of gold price with oil price, stock market and US dollar exchange rates with the updated data set The autoregressive distributed lag bound test (ARDL) and unrestricted error correction models will be used in this study as: (i) they are applied successfully by Hammoudeh et al (2008), Sari et al (2010); (ii) ARDL has many advantages than Engle and Granger (1987) or Johasen (1988) and Johasen and Juselius (1990) that will be mentioned in Chapter 17 The summary of empirical researches listed in Table 3.1 Table 3.1: Summary of empirical studies in Literature review Author Methodology Key variables Period 1995 - 2009 Main Results (year) Narayan et Gregory and Spot and future al (2010) Hansen (1996) prices of gold Gold ↔ Oil in long-run and oil 1990 - 2006 Gold → Oil in Harmmoude ARDL bound test, Spot prices of h et al unrestricted error (2008) correction models, silver, copper; Gold, Oil → US diagostic tests interest rate; dollar index in US dollar long-run oil, gold, short-run; index; some dummy variables Sari et al ARDL bound test, Spot prices of (2010) unrestricted error 1999 - 2007 oil, gold, Oil → Gold in short-run; correction models; silver, Gold → the generalized Palladium, exchange rate forecast error Platinum, variance USD/EUR decompostions, exchange rate the generalized impulse response functions Malliaris et Johansen test for Spot prices of al (2011) long-run gold, oil, euro 2000 - 2007 Gold ↔ Oil in short-run relationship; 18 Pairwise Granger causality; VAR Granger Causalitys; Neural network methodology Kim and Granger causality; Prices of gold, 1970 - 2008 Value of dollar Dilt (2011) Vector oil, value of → gold price, autoregression dollar oil price (VAR); Impulse response functions and variance decomposition Smith Granger causality; Four gold price 1991 - 2001 US stock index (2001) error correction series and six returns → gold model different US returns in short- stock price run indices 1996 - 2007 Stock prices → Gilmore et Johansen and Gold prices, al (2009) Julius (1990); stock price gold prices in Vector error indices of gold short-run correction (VEC) mining model; variance companies, decomposition broad stock and impulse market indices response functions Wang et al Johansen test, Oil price; gold (2010) vector error price; exchange 2006 - 2009 Oil price ↔ stock prices in 19 correction model, rates of US Taiwan; granger causality dollar; stock Gold prices ↔ price indices of oil price; United States, Oil price → Germany, exchange rate; Japan, Taiwan, Gold price → Chian exchange rate; Exchange rate → stock prices in Taiwan Bhunia & Johansen test, Gold price, Das (2012) vector error stock returns in correction model, India 2001 – 2011 Gold price ↔Stock returns Granger causality test Note: ↔ means that bi-directional causality exists between two variables; → means that uni-directional causality exists between two variables 20 CHAPTER 4: RESEARCH METHODOLOGY The chapter is constituted by five sections The first section is research process that gives steps to conduct this research The second section is the model establishment The third one will mention about data collection as well as the sample size The fourth one is steps carried out in data analysis The final section is analysis method that will give details about unit root test and cointegration test chosen 4.1 Research process This study will be conducted through steps described in the Figure 4.1 4.2 Model establishment After reviewing empirical studies, the unrestricted error correction model that successfully applied by Hammoudeh et al (2008), Sari et al (2010) is applied in the paper Since the directions of the relationship among gold price with other variables in the long-run are uncertain, the paper will construct the unrestricted regressions in which each of them is dependent variable 4.3 Data collection The data set utilized for the paper is secondary data It consists of daily time series over January 2007 to December 2012 (1,498 observations for each series) Following the previous researches, oil price, US Dollar Index, S&P 500 Index are considered as the independent variables affecting to gold price Global financial crisis (called as Crisis) occurred from August 2007 up to the end of 2008 (Wikipedia, 2013) will be used as a dummy variable 21 Problem statement - Role of gold in economy, in reserve asset of Central Banks; - Recent policies relating to gold of some Central Banks and of the State Bank of Viet Nam; - Target of Vietnam’s government in stabilizing the gold market Overview of Viet Nam’s gold market - Vietnam’s national gold brand – SJC; - Connection between Vietnam’s gold price and International gold price Literature review - Studies on the relationship between gold and oil prices; - Studies on the relationship between gold price and US dollar exchange rate; - Studies on the relationship between gold price and stock market; - Method and model will be chosen in Thesis Model establishment Unrestricted error correction model (UECM) Data collection - Data of oil price, gold price, US dollar index, S&P500 index; global financial crisis occurred in 2007 and 2008 as dummy variable - Sources: EIA, World Gold Council, Federal Reserves Data analysis Research result - Descriptive analysis, Correlation matrix; - Cointegration analysis: ARDL - Conclusion about the relationships of gold with other variables; the impact of global financial crisis Policy implication Figure 4.1: Research process 22 The sources that data obtained are as following: i Crude oil price, quoted in US dollars/barrel, denoted as OIL The paper chooses West Texas Intermediate (WTI) crude oil spot price as a representative of world oil price obtained from the US’s Energy Information Administration under the website: http://www.eia.gov WTI crude oil price is chosen due to: (1) its quality and prices are often higher than quality and prices of OPEC or Brent crude oil (Wikipedia, 2012); (2) the international price are observed here because United States is the world’s largest consumer of oil while WTI crude oil is the main source of oil for them; ii Gold price, quoted in US dollars/ troy ounces, denoted as GOLD It is the daily average of the London afternoon (pm) fix obtained from the World Gold Council under the website: http://www.gold.org These prices are chosen as representative due to London has been the global center for gold refining and exchange since early 19th century; today, the London gold market is still the largest and the most important gold-trading center in the world iii US dollar index, denoted as USDX It is a measure of the value of the US dollar in comparison to seven other major currencies including Euro, Japanese yen, Canadian dollar, Swiss franc, British Pound, Swedish krona and Australian dollar USDX can be obtained from the United States’ Federal Reserve under the website: http://www.federalreserve.gov iv S&P 500 index, quoted in points, denoted as SPX It is a stock market index based on 500 leading publicly traded American companies’ stock prices It is considered as the best indicator of the US economy or “a bellwether for the US economy” (Wikipedia, 2012) The historical data can be obtained from Federal Reserve Bank of St Louis under the website: http://research.stlouisfed.org It is noted that all variables (dummy variable excluded) are modeled in natural logarithms and takes the first difference This provides stationary time-series data and allows for meaningful independent variables (Sari et al., 2010; Le et al., 2011) 23 4.4 Data analysis Firstly, the paper will carry out the descriptive statistics of all series in level, in log and first different of log level to understand what variable has the highest volatility and average return as well as their distribution forms Secondly, correlation matrix will be built and analyzed to know the correlation relationship among variables Thirdly, Augmented Dickey-Fuller (ADF) test and Philips-Perron (PP) test will be used for stationary and unit root test Finally, the cointegration test will be taken – autoregressive distributed lag (ARDL) bound test To test the equilibrium relationship as well as the causual linkage among variables, many empirical studies used the co-integration tests, such as Engle and Granger (1987) or Johasen (1988) and Johasen and Juselius (1990) Those methods required that all variables to be integrated in the same order of one, that is I(1) So, a pre-testing step for unit roots should be involved to determine the order of integration of variables in models However, in practice, variables are not same order of integration Some variables are stationary at level I(0), while others are stationary at level I(1) or I(2), etc This problem may make the estimating results to be spurious Therefore, to overcome the above problems, the research will use ARDL bound test approach suggested by Pesaran and Shin (1999) and Pesaran et al (2001) This test is based on unrestricted error correction model (UECM) And it is said that ARDL is more advantage than previous mentioned methods: (1) it allows variables to have different orders of integration In particular, it allows some variables to be integrated of order and some of order or mutually cointegrated and it does not require a pre-testing for unit roots; (2) it is more powerful even with a small sample size; (3) It helps to examine the short-run and long-run relationships; helps to determine the causality effects; (4) dummy variables can be included in the test process (Frimpong & Oteng-Abayie, 2006; Hoque & Yusop, 2009) 24 4.5 Analysis method The flow chart of statistical analysis method showed in the Figure 4.2 Stationary test - Augmented Dickey-Fuller (ADF) test; - Phillips-Perron (PP) test Cointegration test - ARDL bound test; - Unrestricted error correction models (UECM) Figure 4.2: Analysis method 4.5.1 Stationary and unit root test Although ARDL bound test approach does not require taking unit root tests, it is critical to conduct those tests to ensure that all variables are either I(0) or I(1) This study will employ the conventional unit root tests widely known as ADF and PP unit root tests Generally, a variable is said to be stationary after differencing d times The variable is integrated of order greater than or equal to is non-stationary However, most economic variables are cointegrated of order 1(Asteriou & Hall, 2007) ADF test based on the choosing of following three regression forms in testing for the existence of a unit root of time series Yt: The difference between the three forms is the deterministic elements α0 and α1T In order to choosing the best equation, it is suggested that we can first plot the data of 25 each variable and observe the graph due to it can indicate the existence or not of the deterministic trend regressors (Binh, 2011) The hypothesis is: H0: δ = (Unit root), H1: δ ≠ Decision rule: If t statistic (t*) is greater than ADF’ critical values (in absolute terms), the null hypothesis (Ho) cannot be rejected It means that unit root exists; If t statistic (t*) is smaller than ADF’ critical values (in absolute terms), the null hypothesis (Ho) can be rejected It means that unit root does not exist However, if problem of serial correlation occurs, Phillip-Perron (PP) test conducted in a similar manner of ADF test will be used alternative 4.5.2 Cointegration test Tải FULL (59 trang): https://bit.ly/3wC8e7O Dự phòng: fb.com/TaiHo123doc.net After the order of integration of each variable is established, the paper will investigate whether the variables under consideration is cointegrated Cointegration implies that causality and long-run equilibrium relationship exists among variables, but the direction of the causal relationship does not indicated For testing the existence of cointegration, the paper uses ARDL bound test This test is computed based on an estimated of unrestricted error correction models (UECM) by Ordinary Least Squares (OLS) estimators (Pesaran et al., 2001) Basically, the bound test developed by Pesaran et al (2001) is the Wald test (Fstatistics version of the bound testing approaches) for the lagged levels variables in the right-hand side of UECM By taking each of the variables in turn as a dependent variable, the research will estimate UECM models as followings: 26 Group without crisis interaction: p p (1) ∆lnGOLDt = ∝0 + ∑i=1 ∝1i ∆lnGOLDt−i + ∑i=1 ∝2i ∆lnOILt−i + ∑pi=1 ∝3i ∆lnUSDX t−i + ∑pi=1 ∝4i ∆lnSPX t−i + ∝5 lnGOLDt−1 + ∝6 lnOILt−1 + ∝7 lnUSDX t−1 + ∝8 lnSPX t−1 + ε1t (2) ∆lnOILt = β0 + ∑ki=1 β1i ∆lnGOLDt−i + ∑ki=1 β2i ∆lnOILt−i + ∑ki=1 β3i ∆lnUSDX t−i + ∑ki=1 β4i ∆lnSPX t−i + β5 lnGOLDt−1 + β6 lnOILt−1 + β7 lnUSDX t−1 + β8 lnSPX t−1 + ε2t (3) ∆lnUSDX t = Υ0 + ∑ri=1 Υ1i ∆lnGOLDt−i + ∑ri=1 Υ2i ∆lnOILt−i + ∑ri=1 Υ3i ∆lnUSDX t−i + ∑ki=1 Υ4i ∆lnSPX t−i + Υ5 lnGOLDt−1 + Υ6 lnOILt−1 + Υ7 lnUSDX t−1 + Υ8 lnSPX t−1 + ε3t (4) ∆lnSPX t = φ0 + ∑ni=1 φ1i ∆lnGOLDt−i + ∑ni=1 φ2i ∆lnOILt−i + ∑ni=1 φ3i ∆lnUSDX t−i + ∑ni=1 φ4i ∆lnSPX t−i + φ5 lnGOLDt−1 + φ6 lnOILt−1 + φ7 lnUSDX t−1 + φ8 lnSPX t−1 + ε4t Group with crisis interaction: p p (1) ∆lnGOLDt = ∝0 + ∑i=1 ∝1i ∆lnGOLDt−i + ∑i=1 ∝2i ∆lnOILt−i + ∑pi=1 ∝3i ∆lnUSDX t−i + ∑pi=1 ∝4i ∆lnSPX t−i + ∝5 lnGOLDt−1 + ∝6 lnOILt−1 + ∝7 lnUSDX t−1 + ∝8 lnSPX t−1 + ∝9 D1G + ∝10 D2O + ∝11 D3U + ∝12 D4S + ε1t Tải FULL (59 trang): https://bit.ly/3wC8e7O Dự phòng: fb.com/TaiHo123doc.net (2) ∆lnOILt = β0 + ∑ki=1 β1i ∆lnGOLDt−i + ∑ki=1 β2i ∆lnOILt−i + ∑ki=1 β3i ∆lnUSDX t−i + ∑ki=1 β4i ∆lnSPX t−i + β5 lnGOLDt−1 + β6 lnOILt−1 + β7 lnUSDX t−1 + β8 lnSPX t−1 + β9 D1G + β10 D2O + β11 D3U + β12 D4S + ε2t (3) ∆lnUSDX t = Υ0 + ∑ri=1 Υ1i ∆lnGOLDt−i + ∑ri=1 Υ2i ∆lnOILt−i + ∑ri=1 Υ3i ∆lnUSDX t−i + ∑ki=1 Υ4i ∆lnSPX t−i + Υ5 lnGOLDt−1 + Υ6 lnOILt−1 + Υ7 lnUSDX t−1 + Υ8 lnSPX t−1 + Υ9 D1G + Υ10 D2O + Υ11 D3U + Υ12 D4S + ε3t 27 ∆lnSPX t = φ0 + ∑ni=1 φ1i ∆lnGOLDt−i + ∑ni=1 φ2i ∆lnOILt−i + (4) ∑ni=1 φ3i ∆lnUSDX t−i + ∑ni=1 φ4i ∆lnSPX t−i + φ5 lnGOLDt−1 + φ6 lnOILt−1 + φ7 lnUSDX t−1 + φ8 lnSPX t−1 + φ9 D1G + φ10 D2O + φ11 D3U + φ12 D4S + ε4t Where: - lnGOLDt is the log of the London gold price; - lnOILt is the log of international crude oil prices which measures the price of West Texas Intermediate (WTI) crude oil; - lnUSDXt is the log of US dollar index; - lnSPXt is the log of S&P 500 index; - D1G = lnGOLDt-1* Crisis; - D2O = lnOILt-1 * Crisis; - D3U = lnUSDXt-1 * Crisis; - D4S = lnSPXt-1 * Crisis; - ∆ is the first difference operator; - p, k, r, n are the lag lengths and determined by the Akaike Information Criterion (AIC) (supporting by Eviews software); - α0, β0, 𝛶0, φ0 are the drift; - αxi, βxi, 𝛶xi, φxi (x = to 4) are the short-run coefficients; - αx, βx, 𝛶x, φx (x = to 8) are the long-run coefficients in Group 1; - αx, βx, 𝛶x, φx (x = to 12) are the long-run coefficients in Group 2; - εxt (x = to 4) are white noise errors; - Crisis = 1, over the period 01/08/2007 – 31/12/2008, elsewhere The null hypothesis of no cointegration in the long run in each equation of Group is that: i Equation 1: H0: α5 = α6 = α7 = α8 = 28 6677748 ... Taiwan; granger causality dollar; stock Gold prices ↔ price indices of oil price; United States, Oil price → Germany, exchange rate; Japan, Taiwan, Gold price → Chian exchange rate; Exchange... HO CHI MINH CITY THE HAGUE VIETNAM THE NETHERLANDS VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS FACTORS AFFECTING THE WORLD’S GOLD PRICE: AN ARDL APPROACH A thesis submitted... obtain the mentioned objectives, the paper will try to answer the following questions: i Do Vietnam’s gold prices and global gold prices correlated? ii What are the factors affecting global gold prices?