国际油价_宏观经济变量与贵金属价格的动态交互影响_英文_朱学红.docx

《国际油价_宏观经济变量与贵金属价格的动态交互影响_英文_朱学红.docx》由会员分享，可在线阅读，更多相关《国际油价_宏观经济变量与贵金属价格的动态交互影响_英文_朱学红.docx（8页珍藏版）》请在淘文阁 - 分享文档赚钱的网站上搜索。

1、 Science ELSEVIER Press Available online at ScienceDirect Trans. Nonferrous Met. Soc. China 25(2015) 669676 Transactions of Nonferrous Metals Society of China Dynamic interacting relationships among international oil prices, macroeconomic variables and precious metal prices Xue-hong ZHU12, Jin-yu

2、CHEN1, Mei-rui ZHONG12 1. School of Business, Central South University, Changsha 410083, China; 2. Institute of Metal Resources Strategy, Central South University, Changsha 410083, China Received 8 January 2014; accepted 4 April 2014 Abstract: From the perspective of long-term and short-term, the me

3、thods of TY causality test, generalized impulse response function, variance decomposition were used to investigate the impacts of international oil prices and macroeconomic variables on Chinese gold, silver and platinum prices, but also the feedback effects of Chinese precious metal prices under thi

4、s impact. The results show that international oil prices play an important role in precious metal price variation both in long-term and short-term, and exchange rate only has an effect in short-term, while interest rate is ineffective in predicting precious metal prices. In addition, precious metal

5、prices have some feedback effects on international oil prices and interest rate in short-term. Key words: international oil price; precious metal price; TY causality test; generalized impulse response function; variance decomposition 1 Introduction Commodity market always attracts attention of resea

6、rchers. Because commodity is closely related to industrial production and it could be efficient sign for inflation. Besides, commodity plays an indispensable role in hedging and multi investment portfolios. Taking gold, silver and platinum as an example, the commodity feature and financial feature o

7、f those precious metals are widely concerned. China is the second net import country of petroleum and the great demand helps to increase oil prices and it could lead to inflation and fluctuation of exchange rate domestically. In this situation, investors turn to precious metals to resist inflation a

8、nd exchange rate risk. Precious metals, to some extent, are regarded as reserve currency which could stimulate production and consumption of precious metals. It should have some effects on exchange rate and interest rate. Intensive need of petroleum and precious metal, to some extent, could drive pr

9、ice linkage among those commodities. Recently, researchers have paid attention to relationship between international oil prices and main macroeconomic variables with prices of precious metals. These literatures could be concluded to four aspects. 1) The relationship between oil prices and precious m

10、etal prices. SARI et al 1 found that oil and silver were shortly related in developed countries and they also displayed the inner relationship mechanism of oil and precious metals 2. ZHANG and WEI 3 advocated that in the sample period, the relation coefficient of prices of gold and oil was 0.9295 an

11、d they had long-term equilibrium relationship. 2) The relationship between main macroeconomic variables and precious metal prices. TULLY and LUCEY 4 found that US dollar was the most important factor and, at most situation, was even the only variable leading to exchange rate fluctuation. By establis

12、hing the error correction model, SJAASTAD 5 got the conclusion that floating exchange rate system was main cause for unsteady international gold price. HAMMOUDEH et al 6 found that the introduction of exchange rate increased direct and indirect impacts on commodity futures in all models. Foundation

13、item: Project (13&ZD169) supported by the Major Program of the National Social Science Foundation, China; Project (13YJAZH149) supported by Research Project in Humanities and Social Sciences Conducted by the Ministry of Education, China; Project (2011ZK2043) supported by the Key Program of the Soft

14、Science Research Project of Hunan Province, China; Project (2015JJ2182) supported by Natural Science Foundation of Hunan Province of China; Project (2009JYJR035) supported by Emergency Project The Study of International Financial Crisis” of Ministry of Education of China Corresponding author: Mei-ru

15、i ZHONG; Tel: +86-15874096280; E-mail: DOI: 10.1016/S1003-6326(15)63651-2 670 Xue-hong ZHU, et al/Trans. Nonferrous Met. Soc. China 25(2015) 669-676 3) The relationship between oil prices and macro- economic variables. The research of AMANO and NORDEN 7 presented that in the relationship between oi

16、l price and real exchange rate, oil price took the dominant position. And interest rate was significant variable linking to oil market, exchange rate and currency policy. Oil price could offer some information to the market participants. SJAASTAD and SCACCIAVILLANI 8 indicated that increasing oil pr

17、ice resulted in depreciation of US dollar within net oil exporting countries. TANG and LUO 9 discovered that shocks of oil price lead to appreciation of real exchange rate in China. 4) The relationship among precious metal prices. CINER 10 discussed the absence of the long-term relationship between

18、gold and silver in Tokyo Commodity Exchange in 1990s. LUCEY and TULLY 11 considered that the steady relationship between gold and silver could be sustained and even the relationship was weak at this moment. By GARCH models and EG ARCH models, MORALES 12 studied volatility spillover effect among gold

19、, silver, platinum and palladium, and there was evidence that the relationship was bidirectional. But there was little evidence that other precious metal markets could influence gold market. Besides, some researchers are trying to study the relationship between oil prices and macroeconomic variables

20、 with precious metal prices in an integrated frame. SARI et al 2 found that there was weak long-term equilibrium relationship among precious metal prices, oil prices and exchange rate. By VAR model, HUANG et al 13 analyzed impacts of US dollar and oil price on Chinese copper, gold and silver. The re

21、sult showed that the relationship between US dollar, gold and silver determined prices of Chinese gold and silver. Based on weekly data, LI and FU 14 established VAR-DCC-MVGARCH model to estimate the dynamic relationship among oil, gold, interest rate, exchange rate and stock market in China. Accord

22、ing to the above description, researchers around the world have not received an agreement about relationship among oil prices, macroeconomic variables and precious metal prices. At this moment the existing literatures are lack of synthesis and are aimed at developed countries. With economical and fi

23、nancial development in emerging countries, those emerging capital markets could present more global shocks. Besides, economy situations are different between developed and developing countries, which could lead to completely different conclusions. As a result, constructing a new frame directing to e

24、merging market, such as China, to discuss dynamic relationship among petroleum, macroeconomic variables and precious metals seems very necessary. Based on TY causality test, generalized impulse response function and variance decomposition, bidirectional dynamic characteristics of international oil p

25、rices, interest rate, exchange rate and Chinese precious metal prices were studied. Main innovations include three aspects. 1) Information loss could be avoided by without considering cointegration test of TY causality test and rankings of variables in VAR model of generalized impulse response funct

26、ion, so interactive relationship among variables could be precisely described. 2) From short-term and long-term perspective, the dynamic relationship among global oil prices, interest rate, exchange rate and Chinese precious metal prices is reflected in time and in section-cross series. 3) Some new

27、issues are added in this work, such as the relationship between oil and interest rate, the relationship between interest rate and precious metals. 2 Methodology 2.1 TY causality test In order to determine the long-term relationship among variables, this work is based on TY causality test. Unlike oth

28、er common causality tests, the main advantage of the TY procedure is that in this procedure there is no need to test for cointegration. Hence, a likely pretest bias is avoided. Furthermore, the TY procedure allows us to run a VAR in levels, regardless of whether the series have the same order of int

29、egration or not. Therefore, there is no information loss due to differencing and the procedure is more flexible in considering arbitrary levels of integration. According to TOD A and YAMAMOTO 15, the following VAR representation for Yt is considered: Yt =0 + 0 +- + OkYt_k +ut (1) where =(为 ,， .)； ar

30、e coefficients to be estimated; c is a vector of constants; t is linear time trend; k is the optimal lag length; and ut is a vector well-behaved disturbances. In this model, the null hypothesis is that there is no causality, the null hypothesis can be expressed as elements in coefficient matrixes th

31、at are all equal to zero. The TY procedure starts with determining the maximum integration order d) for the series in concern. Then, the optimum lag length (k) is determined via some information criteria. If the augmented vector autoregression VAR( +) satisfies the common assumptions, a Wald test on

32、 the joint significance of the first k lags of each variable constitutes a long-run Granger causality test. The Wald test statistic follows an asymptotic Chi-square distribution with k degrees of freedom. Two problems need to be solved to process TY causality test: the maximum integration order d) f

33、or the Xue-hong ZHU, et al/Trans. Nonferrous Met. Soc. China 25(2015) 669-676 series and the optimum lag length (k) for VAR model. We employ ADF unit root test to assess the order of integration, and use LR statistic, Akaike Information Criterion or Schwartz Criterion to determine the lag order for

34、VAR model. 2.2 Generalized impulse response function In order to study the short-term dynamic impact reaction among international oil prices, interest rate, exchange rate and precious metal prices, this research is also based on GIRF, which is presented by KOOP et al 16. The basic thought of GIRF is

35、 expressed as Gn,Sj,) = Ext+n Isjt = Sj-Eix Iat_x) (2) 671 This is the result of the influence of the jh disturbing term on the /th variable from the infinite past to the current time point, which is assessed by the variance. Here, it is assumed that covariance matrix of vector of disturbing term is

36、 diagonal matrix, so the variance of yt is the k diagonal matrix of the above variance: var(j) = 2fp2 j (8) y=i U= J The variance of yt can be dissembled into k irrelevant influences. Therefore, in order to measure the degree of influence of each disturbing term on variance of yh the following measu

37、rement is defined: where Sj represents the shocks from the /th variable, and ot_x represents the set of all available information when shocks occur. Further, assuming that st has a multivariate normal distribution, it is now easily seen that ir)2 ( ) /7=0 /7=0 (9) E(St 1 Sj, = Sj) = (Tly(T2y-0 )， 0i

38、,1iodssuods9Di Xue-hong ZHU, et al/Trans. Nonferrous Met. Soc. China 25(2015) 669-676 675 Table 5 Variance decomposition results of gold and other variables Item Effect of other Effect of lgold on variables on lgold/% other variables/% lbrent 6.473 0.794 lsilver 0.054 36.556 lplatinum 0.608 36.070 l

39、bond 0.081 0.115 ler 0.002 4.020 Table 6 Variance decomposition results of silver and other variables Effect of other variables Item on lsilver/% Effect of lsilver on other variables/% lbrent 7.081 0.136 lgold 36.556 0.054 lplatinum 0.011 0.808 lbond 0.006 0.406 ler 0.046 0.027 Table 7 Variance deco

40、mposition results of platinum and other variables Effect of other variables Effect of lplatinum on on lplatinum/% other variables/% lbrent 15.134 0.567 lgold 36.070 0.608 lsilver 0.808 0.011 lbond 0.011 0.131 ler 0.023 0.768 oil prices, which is up to over 6% and the impact of international oil pric

41、es on forecast error variance of platinum even reaches more than 15%, while the variance of each precious metal explained by interest rate and exchange rate is quite small, which is below 0.1%. This analysis results further validate that domestic precious metal prices are mainly affected by internat

42、ional factors such as international oil prices. Domestic macroeconomic variables have failed to predict the price movements of precious metals. In contrast, each precious metal does not significantly explain the variance of international oil prices, interest rate and exchange rate, but in terms of t

43、he interaction between the precious metals, the impact of gold on forecast error variances of silver and platinum is great, which are respectively 36.5% and 36%. This is because gold has larger trading volume, higher liquidity and wider recognition in China, which results in a significant impact on

44、silver, platinum and other precious metal prices. 4 Conclusions 1) Both in long and short term, international oil price is an important factor for precious metal prices volatility, exchange rate will have a negative impact on precious metal prices in the short term, while interest rate fails to pred

45、ict the precious metal prices. 2) Precious metal prices provide some feedback effects on oil price and interest rate in short term, which precious metal prices will have a positive effect on international oil prices and impose a negative impact on interest rate. 3) International oil prices play an i

46、mportant role in explaining the forecast error variance of precious metal prices volatility. However, the variance of each precious metal explained by interest rate and exchange rate is quite small. This result reminds us that on one hand, great attention should be paid to prevent imported inflation

47、 caused by rising international oil prices; on the other hand, the function of domestic macroeconomic variables predicting precious metal prices has not been fully realized. 4) Overall, in terms of the interaction between precious metals, gold plays a vital role in explaining the forecast error vari

48、ance of other precious metals. It shows that in China, gold as an investment and monetary asset bears more advantages than other precious metals and fits to be a major haven for funds. References 1 SARI R, HAMMOUDEH S, EWING B T. Dynamic relationships between oil and metal commodity futures prices J

49、. Geopolitics of Energy, 2007, 29(7): 2-13. 2 SARI R, HAMMOUDEH S, SOYTAS U. Dynamics of oil price, precious metal prices, and exchange rate J. Energy Economics, 2010,32: 351-362. 3 ZHANG Y J, WEI Y M. The crude oil market and the gold market: Evidence for cointegration, causality and price discovery J. Resources Policy, 2010, 35: 168-177. 4 TULLY E, LUCEY B. A power GARCH examination of the gold market J. Research in Internationa