This research aims at testing the correlation between the price of gold and many macroeconomic variables in Vietnam. Multiple Linear Regression is employed to determined significant relationship between dependent and independent variables, covering the data collected from January 2010 to June 2016.
Chính sách & thị trường tài - tiền tệ Các yếu tố tác động đến giá vàng thị trường Việt nam Trịnh Anh Khoa Nghiên cứu thực nhằm mục tiêu kiểm định tương quan giá vàng số yếu tố kinh tế vĩ mô Việt Nam Trên sở liệu thu thập từ tháng 01/2010 đến tháng 6/2016, mơ hình hồi quy tuyến tính đa biến ứng dụng để kiểm định mối tương quan giá vàng với biến độc lập mơ hình nghiên cứu Các biến độc lập đưa vào mơ hình nghiên cứu bao gồm: giá dầu thô, CPI, tỷ giá USD, số VN- Index số HNX- Index Kết nghiên cứu cho thấy: số VN- Index HNX- Index có tương quan nghịch với giá vàng giá dầu thơ lại có tương quan thuận với giá vàng Tuy nhiên, kết nghiên cứu cho thấy tương quan CPI tỷ giá USD với giá vàng khơng có ý nghĩa thống kê Từ khóa: CPI, giá dầu thô, yếu tố tác động, giá vàng, tỷ giá USD Introduction old is one of the most important metals in almost all economies Gold as an asset has a hybrid nature: It is not only a commodity used in many industries, but also has maintained throughout history as a means of exchange and a store of value (Sujit and Kumar, 2011) With the second function, gold has been becoming a deliriously important channel of investment instead of securities The demand of gold has been on roll during the last decade as well as non- monetary requirement THÁNG 1&2.2017 - SỐ 176+177 All the countries in the world use gold as their means of transaction in international trade Gold is now being used as an alternative for dollar since its collapse In general, during the time of political and economic uncertainty, investors tend to purchase gold and gold related instruments as a store of value, a diversification tool to protect themselves from stock and currency shocks (Kiohos and Sariannidis, 2010) In many developing countries, where informations on financial markets consisting of stock market and commercial banks are still in deficiency, gold becomes an attractive investment channel However, as many other yielding assets, gold causes some fluctuations that can harm the returns of investors In world view, there are a lot of factors that could affecting the price of gold In Vietnam, the awareness of people or investors on determinants of gold prices has grown in the past few years This study aims to indentify factors affecting the price of gold in Vietnam The results will be valuable for policy makers, investors and people in making their decisions 67 Figure 1: Volidity of gold prices and economic factors (a) Gold and Brent Crude Oil price from 1987 to 2012 (b) Growth of Gold prices and USA CPI from 1955 to 2005 (c) Gold prices and USD from 1995 to 2015 Source: CME, EIA Source: usagold.com Source: macrotrend.com Literature review There are many factors that contribute to the formation of gold prices History has shown that accompany with the changes of 68 gold prices were the volatility of macroeconomic factors including CPI, oil prices, exchange rate and so on Some existing experimental studies look into comovement, co-integration and relationship between macroeconomic factors and gold prices on the global scope USD exchange rate As far as the exchange rate mar- SỐ 176+177 - THÁNG 1&2.2017 ket is concerned, it has long been thought that gold was a good protection from fluctuations of the USD, the main currency traded in the world Jaffe (1989), Tully et al (2007) and Nair et al (2015) found that there was a correlation between gold prices and exchange value of USD dollar By using factor analysis for the period 1990 – 2011, Elfakhani, Baalbaki, and Rizk (2009) explored that the value of the US dollar, which indicated in the study as USD index, negatively impacted on the gold prices In the same view, finding of Ibrahim, Kamaruddin and Hasan (2014) by using Multiple Linear Regression Model for analysing the data collected from 2003 until 2012 provided an evidence suggesting the negative correlation between gold prices and exchange rate Furthermore, Pravit (2009) presented the research result using Multiple Regression and Auto Regressive Integrated Moving Average which indicated that US dollars had negative effects on the changes of Thai gold market meanwhile Australian Dollars, Japanese Yen, Canadian Dollars and EU Ponds had positive impacts on the price of gold Oil price Regarding the effects of oil prices to gold market, they influence the cost of production of goods and services, a fact that predicts the future of the other industries and also affects the profit margins The study by Zang et al (2010) analyzed the cointegration relationship between gold and crude oil prices The study found that the change of crude oil prices caused the positive volatility of gold prices More- THÁNG 1&2.2017 - SỐ 176+177 over, the oil and gold prices are considered interconnected in Thailand and Malaysia through the studies of Pravit (2009) and Ibrahim, Kamaruddin and Hasan (2014) In these research results, the increase of crude oil prices provoke the demand of gold, hence, its prices are anticipated to increase The research of Klohos and Suriannidis (2010) to explore in the shortrun effects of energy (crude oil) and financial (equity, currency and bond) markets on the gold market has been concerned above also gave the proof about effect of crude oil prices on gold prices The results suggested that the energy market positively influences the gold market Inflation or CPI Inflation has been considered as a driving factor for the value of gold Because the main purpose of holding gold is to store value, the price of gold was expected to go up with inflation (Elfakhani, Baalbaki, and Rizk, 2009) Moreover, the gold prices is often counted as a cause and effect relationship to inflation (Klohos and Suriannidis, 2010) By using Pooled Ordinary Least Squares (POLS) for data employed across a 14 years period from year 2000 until 2013, Shukur et al (2015) confirmed the highly positive contribution to gold prices of inflation rate in Malaysia market Another study in Malaysia carried out by Ibrahim, Kamaruddin and Hasan (2014) showed a negatively significant relationship between inflation rates and gold prices Nonetheless, Abken (1980), Sjaastad (2008) and Toraman, Basarir and Bayramoglu (2011), found that inflation does not have any significant impact on the gold prices Noticeably, the figure used to measure inflation rates in these researches is CPI (Consumer Price Index) Stock activities The effect of stock markets on the price of gold could be regarded particularly important in the process of defining portfolios, evaluating, tracking portfolio performance A truthworthy explanation for this relationship is that the gold market functions as a mobilization factor of hedge to protect investors from portfolio and geopolitical risks (Klohos and Suriannidis, 2010) “Investment instruments on the stock market may be competitive substitutes for putting capital in gold, and investor disinterest in gold may have something to with a lack of promotion” (Elfakhani, Baalbaki, and Rizk, 2009) Also according to the researches of Elfakhani, Baalbaki, and Rizk (2009), Klohos and Suriannidis (2010), they found that the equity markets through S&P 500 exert negatively contribution to the gold market In other words, an increase in S&P 500 often causes a decrease in gold prices Sujit and Rumar (2011) came to similar conclusions claiming that the increase of S&P 500 will drives the price of gold down Nevertheless, Kothari and Gulati (2015) argued that stock indecies positively correlate to gold prices as a result from the study in Indian stock and gold markets Whereas, Najafabadi, Qazvini and Ofoghi (2012) concluded that there is no significant direct relationship between gold prices and the TSE index, but the TSE is indirectly influenced by 69 gold price through other factors This study aims to test the correlation between factors concluding exchange rate, crude oil prices, Consumer Price Index and stock activities and the gold prices in Vietnam Data and Research methodology For this research, the second data for years and a half starting from January 2010 to June 2016 has been used The data of CPI and USD exchange rate were gathered from Monthly Statistic Reports of General Statistic Office Then, the impacts of stock activities were measured by stock indicies concluding VNIndex and HNX-Index as two important figures standing for the volidity on stock market quoted on Ho Chi Minh Stock Exchange and Ha Noi Stock Exchange The price of crude oil and gold were used in the research are available on many major statistic websites such as: sjc.com, doji.com, giavang.com, mof.gov.vn and petrolimex.com Multiple Linear Regression model has been used as statistical tool for this study The model was used to test the impacts of five factors consisting of CPI, USD exchange rate, crude oil prices, VN-Index and HNX- Index towards the dependent variable is gold prices Multiple Linear Regression analysis give informations about how much of the variance of the dependent variable is explained by independent variables “This model is more appropriate to be used since it can explain the correlation between the dependent and independent variables much better” (Ibrahim, Kamaruddin and Hasan, 2014) The model for this study is modified from that of Ibrahim, Kamaruddin and Hasan (2014) and Pravit (2009) for the determination of factors affecting the price of gold The model is as follows: LnGP = α + β1CPI + β2LnER + β3LnOP + β4LnVNI + β5LnHNX + ε Where, LnGP : Logarit of Gold prices Price of gold measured by Vietnamese unit in million VND LnCPI : Consumer Price Index Monthly CPI calculated by General Statistic Office LnER: Logarit of USD exchange rate Monthly USD Index calculated by General Statistic Office LnOP : Logarit of the price of crude oil Price of crude oil in USD LnVNI : Logarit of the value of VN-Index LnHNX : Logarit of the value of HNX-Index βi : The coefficient for the dependent variable ε : Error terms Results and Discussion Table provides an overview of the dependent and independent variables calculated from selected sample The total observation included in sample is 78 observations As can be seen from the table, there was a delirious fluctuation in values of variables as the standard deviation value of each factor is too high The dependent variable is gold prices which the mean and standard deviation is 37.21 and 5.45 respectively A wide gap between maximum and minimum value illustrates that there is a strong volidity of gold prices in the research’s period Among five independent factors, HNX-Index, crude oil prices and Consumer Price Index are three determinants with the most significant difference in maximum and minimum values in reverse order The first independent variable is CPI which the mean and standard deviation is 0.6 and 0.75 respectively Particularly, the maximum value is nearly nine times higher than the minimum value The maximum value of crude oil prices reaches 112.8 nearly three times higher than Table Characteristics of variables GP CPI OP HNX ER VNI 70 Units Million VND % USD Point % Point Mean 37.21 0.60 81.03 85.11 101.01 502.44 Max 48.34 3.32 112.80 175.98 101.09 632.27 Min Standard dev 26.24 5.45 -0.44 0.75 29.80 22.35 51.23 29.15 100.98 1.00 360.37 72.86 Source: Investigated by author 2016 SỐ 176+177 - THÁNG 1&2.2017 Table Pearson correlation matrix CPI 1.000 0.515 0.344 0.094 -0.407 CPI LnER LnOP LnHNX LnVNI LnER 0.515 1.000 0.228 0.421 -0.408 LnOP 0.344 0.228 1.000 -0.167 -0.514 LnHNX LnVNI 0.094 -0.407 0.421 -0.408 -0.167 -0.514 1.000 0.283 0.283 1.000 Source: Results extracted from SPSS Table Variance analysis Regression Residual Total Sum of Squares 1.447 0.265 1.711 df 72 77 Mean Square 0.289 0.004 F 78.711 Sig 0.000a a Predictors: (Constant), CPI, LnHNX, LnOP, LnVNI, LnER b Dependent Variable: LnGP Source: Results extracted from SPSS Table Spearman correlation test CPI Sig two-tailed 0.114 the minimum value which is just 29.8 In addition, the standard deviation of this determinant is approximately seventy five percent of the minimum value The third variable is HNX which the mean and standard deviation is 85.11 and 29.15 respectively The maximum value is approximately 2.5 times higher than the minimum value and the standard deviation is really high comparing with the minimum value The two last variables are considered change modestly in comparision with three factors above The mean and standard deviation of EX is respectively 101.01 and However, that of VNX-Index should be regarded and the maximum value of this factor is approximately double the minimum value Taking all the above statistic analysis, there is a fluctuation in gold prices in Vietnam through THÁNG 1&2.2017 - SỐ 176+177 LnER 0.383 LnVNI 0.150 the period of research Accompany with the changes of the gold prices is the volidities on many other markets including energy and stock market, but different in level The Least Square method was used then to test if there was a relationship between gold prices and these factors or just a random phenomenon To be considered as an appropriate model for examing the correltion between dependent and independent variables, the model should satisfy three conditions are there is no multi-collinearity and autocorrelation between independent variables and the variance of residual is stable Some tests were carried out to make decision if the model is fixed or not The table below shows the results from Pearson correlate test As can be seen from the table that the correlation values of predictors are less than 0.6 More- LnOP LnHNX 0.091 0.299 Source: Results extracted from SPSS over, the values of collinearity statistics (VIF) are less than 10 (table 5) Hence, there is no multi-collinearity among independent variables The F test was used to estimate the appropriation of Multiple Linear Regression The significant value of F test is below 0.05 (table 3) Therefore, the coefficient of the independent variables are different from zero In other words, the model based on thesis is suitable to practical situations Table show the results of Spearman correlation test between independent variables and residual The values of significant are over 0.05, thus, the residual of model is stable Below is the table that shows the summary of the model The adjusted R square was 0.835 indicates that 83.5% of the variance in gold prices was significantly 71 Table Summary of model Adjusted R Square Std Error of the Estimate 0.835 0.061 Change Statistics F Change df1 df2 78.711 72 Sig 0.000 DW 2.009 a Predictors: (Constant), CPI, LnHNX, LnOP, LnVNI, LnER b Dependent Variable: LnGP Source: Results extracted from SPSS explained by predictors The remaining 16.5% was explained by other factors that not included in the model Durbin-Watson value (DW) is 2.009 supporting that there was no autocorrelation between independent variables According to the results presented in table 6, there are three factors that contribute to the changes of gold prices in Vietnam concluding crude oil prices, VN-Index and HNX-Index The results also shown that there are strong correlations between these factors and the price of gold since the value of adjusted R square is in excess of 80% Based on analysis in table 5, this study found that there is significant correlation between gold prices and stock activities as standardized coefficients beta of stock indicies in the model are too high in a comparision with that of variable crude oil prices The t-statistic value of HNX is 12.457, six times higher than and the significant value stand at 0.000 on 0.05 level While that of VN-Index are 3.544 and 0.001 respectively The standardized coefficients beta of these predictors are negative numbers that mean any increase in HNX-Index or VN-Index leads to a decrease in the price of gold Gold is different from other assets because of its high liquidity Investors use simply risk management strategy is diversifying their portfolio including gold when the stock market flutuates This finding is similar to recent studies and it contributes a proof of negative relationship between gold and stock market in Vietnam The second noticeable result is the positive correlation between crude oil prices and gold prices with t-statistic value reaches 2.7 and the significant stand at 0.09 on 0.05 level The standardized coefficients beta is 0.066 imply- ing that when the price of crude oil goes up, the gold prices will increase As had been concerned above, many researches found that crude oil prices have positive effects on gold prices The oil and gold prices are often considered interconnected through their links to inflation Higher inflation seems to generate more demand of holding gold and gold has been considered as a good inflation hedge tool The increase of crude oil prices often affects inflationary pressures, hence, gold is regarded as a more secure way for storing, the demand of gold and its price is anticipated to increase (Pravit, 2009; Klohos and Suriannidis, 2010) Lastly, based on the results between two other determinants and gold prices found that there are no significant correlations among these variables The significance of USD index and CPI are 0.86 and 0.565 respectively Table Final fiding of Multiple Linear Regression Unstandardized Coefficients B Std Error (Constant) 6.527 0.429 LnER -0.048 0.270 LnVNI -0.238 0.067 LnOP 0.066 0.024 LnHNX -0.395 0.032 CPI 0.649 1.124 Dependent Variable: LnGP 72 Standardized Coefficients Beta -0.012 -0.235 0.148 -0.763 0.033 t Sig 15.231 -0.177 -3.544 2.700 -12.457 0.577 0.000 0.860 0.001 0.009 0.000 0.565 Collinearity Statistics VIF 2.198 2.055 1.408 1.746 1.502 Source: Results extracted from SPSS SỐ 176+177 - THÁNG 1&2.2017 implying that the change of USD index or CPI have no impacts on the fluctuation of gold prices A possible explaination for this results is that there are many restrictions in exchange USD among individuals in Vietnam, hence, when the one’s price goes up they can not buy more another for storage, as a result, there is no correlation between these two factors Regarding CPI, this result is similar to the findings of Abken (1980), Sjaastad (2008) and Toraman et al (2011) According to these authors, gold is not widely used as a store of value against inflation, but people prefer to hedge their wealth by investing in other assets such as houses, stock and so on This is a truthworthy explaination for case of Vietnam Conclusions This research examined the effects of macroeconomic factors on the price of gold in Vietnam using Multiple Linear Regression model More specifically, the study aims to test the influence of crude oil prices, CPI, USD exchange rate, VN-Index and HNX-Index on the gold prices based on the data collected January 2010 to June 2016 As an overall conclusion from the empirical test results, there are significant positive correlations between gold prices with oil prices Meanwhile, the price of gold has negative correlation with stock indicies On the other hand, the research results also found that CPI and USD Index have no role in the fluctuation of gold prices The explanation of this empirical fact is due to a number of specific conditions on the foreign currency market in Vietnam The study generally achieves the objective where to determine the effect of crude oil prices, CPI, USD exchange rate, VN-Index and HNX-Index on the gold prices Even though, there are many factors that can affect the price of gold but has not been included in the model The researcher recommends that further researches should conclude more macrovariables Also, the USD exchange rate should be considered in further researches to reaffirm this research results Moreover, the research will become more reliable if the data was collected in a longer period as 20 years or more ■ Tài liệu tham khảo Abken, P A (1980) ‘The Economics of Gold prices Movements’, Federal Reserve Bank of Richmond (Economic Review), No September, p.3-13 Elfakhani, S., Baalbaki I B., and Rizk H (2009) ‘Gold price determinants: empirical analysis and implications’, Journal for International Business and Entrepreneurship Development, Vol 4, p.161-178 Ibrahim, S N., Kamaruddin, N I and Hasan, R (2014) ‘The determinants of gold prices in Malaysia’, Journal of Advanced Management, Vol 2, p.38-41 Jaffe, J F (1989) ‘Gold and Gold Stocks as Investments for Institutional Portfolios’, Financial Analysts Journal, Vol 45 (2), p.53-59 Kiohos, A and Sariannidis, N (2010) ‘Determinants of the asymmetric gold market’, Journal of Investment Management and Fiancial Innovations, Vol 7, p.26-33 Kothari, A and Gulati, D (2015) ‘Investment in gold and stock market: an analytical comparision’, Pacific Business Review International, Vol 7, p.65-68 Nair, G K., Choudhary, N and Purohit, H (2015) ‘The relationship between gold prices and exchange value of US Dollar in India’, Emerging Markets Journal, Vol 5, p.17-25 Najafabadi, A T P., Qazvini, M and Ofoghi, R (2012) ‘The impact of oil gold prices’ shock on Tehran stock exchange: A Copula approach’, Iranian Journal of Economic Studies, Vol 1, 23 - 47 Pravit, K (2009) ‘Forecasting Thai Gold Prices’, http://www.wbiconpro.com/3-Pravit-.pdf 10 Sjaastad, L.A (2008) ‘The Price of Gold and the Exchange Rates: Once Again’, Resource Policy, vol 33 (2), p.118-124 11 Sujit, K S and Kumar, B R (2011) ‘Study on dynamic relationship among gold price, oil price, exchange rate and stock market returns’, International Journal of Applied Business and Economic Research, Vol 9, p.145-165 12 Toraman, C., Basarir, C and Bayramoglu, M (2011) ‘Determination of factors affecting the price of gold: a study of MGARCH model’, Business and Economics Research Journal, Vol 2, p.37-50 13 Tully, E and Lucey, B M (2006) ‘A power GARCH examination of the gold market’, Research in International Business and Finance, Vol 21, p.316-325 14 Zakira, H., Shukur, N A., and Affandi, S (2015) ‘Factors affecting the price of gold in Malaysia’, Journal of Basic and Applied Scientific Research, Vol (7), p.41-46 15 Zhang, Yue-Jun, Wei Yi-Ming (2010) ‘The Crude Oil Market and the Gold Market: Evidence for cointegration, Causality and Price Discovery’, Resource Policy, Vol 35, p.168-177 THÁNG 1&2.2017 - SỐ 176+177 73 Thông tin tác giả Trịnh Anh Khoa Đơn vị công tác: Khoa Kinh tế, Cao đẳng Cộng đồng Sóc Trăng Tạp chí tiêu biểu có viết đăng tải: Tạp chí Cơng Thương, Tạp chí Khoa học Đại học Cần Thơ, Tạp chí Nhân lực Khoa học Xã Hội Email: takhoastcc@gmail.com Summary The determinants of gold prices in Vietnam This research aims at testing the correlation between the price of gold and many macroeconomic variables in Vietnam Multiple Linear Regression is employed to determined significant relationship between dependent and independent variables, covering the data collected from January 2010 to June 2016 Five independent variables that affect the gold prices are crude oil prices, CPI, USD exchange rate, VN-Index and HNX-Index The results shows that there is negatively significant relationship between VN-Index, HNX-Index and gold prices, meanwhile crude oil price is positively significant However, CPI and USD exchange rate have no role in affecting the price of gold Keywords: CPI, crude oil prices, determinant, gold prices, USD exchange rate Khoa Anh Trinh, Sai Gon Commercial Bank, Bac Ninh Branch Received: 22 August 2016 / Accepted: 15 October 2016 trang 42 The Granger causality test is conducted to answer for the questions whether inflation is useful in forecasting inflation uncertainty and whether inflation uncertainty is useful in forecasting inflation In order to that test, inflation uncertainty firstly must be checked stationarity Similarly, ADF and PP test also are used with results of t-statistic are14.76666 and-14.76636 respectively, which are smaller than the critical value of 1% level Hence, inflation uncertainty is stationary at 1% significance level It is easily realized that the p-value of the first hypothesis is less than 0.01, or the fist null hypothesis is rejected at 1% level This result suggests a positive causal relation from inflation to inflation uncertainty In the second hypothesis, the p-value is larger than 0.01, so it can’t be rejected It means that there doesn’t exist causal relation from inflation uncertainty to inflation This re- 74 sult is completely consistent with Friedman hypothesis Conclusion The research applied AR(1) and EGARCH(1,1) to determine inflation uncertainty and test the relationship between inflation and inflation uncertainty in Vietnam over the period from January 1995 to June 2016 Then, Granger causality was used to test causality between inflation and inflation uncertainty The empirical results provided concrete evidence that there is positive relationship between inflation and inflation uncertainty in Vietnam, and inflation causes inflation uncertainty In order to stabilize inflation as well as minimize negative impacts of inflation uncertainty in Vietnam, this may be archived through following Firstly, the central bank should continue to prioritize the inflation objec- tive and synchronously implement the inflation targeting policy Secondly, the central bank should have quick policy responses to inflation developments, thereby reducing inflation uncertainty both in the short and long run Thirdly, inflation in Vietnam much stems from inflation expectations, so government should establish suitable channels to share information of domestic inflation with the general public in order to reduce inflation expectations, thereby reduce inflation uncertainty ■ SỐ 176+177 - THÁNG 1&2.2017 ... price of gold Oil price Regarding the effects of oil prices to gold market, they influence the cost of production of goods and services, a fact that predicts the future of the other industries... takhoastcc@gmail.com Summary The determinants of gold prices in Vietnam This research aims at testing the correlation between the price of gold and many macroeconomic variables in Vietnam Multiple Linear Regression... variables According to the results presented in table 6, there are three factors that contribute to the changes of gold prices in Vietnam concluding crude oil prices, VN-Index and HNX-Index The results