我国黄金现货、ETF、期货市场动态联动及波动溢出效应  

Dynamic Linkage and Volatility Spillover Effect of Gold Spot, ETF and Futures Markets in China

作　　者：曹栋[1] 徐静静 李汶蔚 赵婕 Dong CAO;Jingjing XU;Wenwei LI;Jie ZHAO(School of Economics and Management,Xidian University,Xi'an 710126,China)

机构地区：[1]西安电子科技大学经济与管理学院,西安710126

出　　处：《计量经济学报》2023年第4期1154-1175,共22页China Journal of Econometrics

基　　金：陕西省社会科学基金(2022YA14);中央高校基本科研业务费(KYFZ23008);西安电子科技大学杭州研究院2023年度概念验证基金(GNYZ2023ZC0503);西安市社会科学规划基金(23JX66)。

摘　　要：黄金现货、ETF、期货等市场在维护一国经济稳定、增强国家信用、对冲金融市场波动等方面一直发挥着核心作用.最近几年,黄金价格历经多次剧烈波动.黄金现货、ETF、期货之间观察到了不同步不协同现象,加大了市场不确定性和风险.本文选取2015年至2021年黄金现货合约、华安黄金ETF、黄金期货主力合约每日收盘数据,探讨黄金现货、ETF、期货市场的动态联动及波动溢出效应.首先,构建MS-GARCH模型,研究三市场波动及区制转变关系;然后借助DCC-GARCH模型进一步探究三个黄金市场彼此之间的动态联动关系;最后使用溢出指数模型测度三市场间的波动溢出效应.结果表明:1)三市场一体化程度较高.DCC-GARCH模型表明:三市场收益率动态相关性绝大部分时间都维持在了0.9左右的水平.2)不同市场之间的动态联动性具有显著区别.DCC-GARCH模型表明:黄金现货与期货市场收益率之间的相关系数波动更为剧烈更为频繁.溢出指数模型表明:黄金现货市场在多数时期呈现为负向净溢出,绝大部分时间为溢出的接收者,黄金ETF、期货市场收益率多数时期为正向溢出,即绝大部分时间为溢出的传播者.3)样本期内,三市场波动溢出具有动态特性.MS-GARCH模型平滑概率图发现2019年后三市场收益率在高、低波动状态的持续时间都比较短,两种状态的交替更加频繁.DCC-GARCH模型表明三市场收益率相关系数在2015年、2018年后都有显著下降.溢出指数模型表明;三市场的总溢出指数在2019年出现较大波动,一度下探到55%;2019年对外溢出指数(TO)和接收溢出(FROM)均出现了显著的下降,黄金现货市场降幅最大,分别下降到43%、50%;2019年黄金现货、ETF市场收益率的净溢出指数波动较大,也观察到了黄金现货到ETF、现货到期货的净成对溢出指数的大幅波动.Gold spot,ETF and futures markets play important roles in maintaining a country's economic stability,enhancing national credit and hedging financial market fluctuations.In recent years,the price of gold has fluctuated greatly,and the phenomenon of unsynchronized and uncoordinated gold spot,ETF and futures has been observed,which has increased market uncertainty and risks.This paper selects the daily closing data of gold spot contracts,Huaan gold ETF and gold futures main contracts from 2015 to 2021 to discuss the dynamic linkage and volatility spillover effects among gold spot,ETF and futures markets.Firstly,the MS-GARCH model is built to study the relationship between price fluctuation and regional system transformation in the three markets.Then through DCC-GARCH model to explore the dynamic linkage between gold futures,spot and ETF markets.Furthermore,the spillover index model is used to measure the volatility spillover effect among the three markets.The results show that:1)The integration degree of the three markets is high.DCC-GARCH model shows that the dynamic correlations of the three market returns are maintained at about 0.9 for most of the time.2)There are significant differences in dynamic linkage among different markets.DCC-GARCH model shows that the correlation coefficient between gold spot and futures market returns fluctuates more violently and frequently.Spillover index model shows that the gold spot market presents negative net spillover in most periods and is the receiver of spillover in most periods.While the return rate of gold ETF and futures market is positive spillover in most periods,that is,it is the disseminator of spillover in most periods.3)During the sample period,the volatility spillovers of the three markets are dynamic.The smoothing probability diagram of MS-GARCH model shows that after 2019,the duration of the three market returns in the high and low volatility states is relatively short,and the alternation of the two states is more frequent.DCC-GARCH model shows that the correlati

关 键 词：黄金现货市场 ETF市场 期货市场 动态联动 MS-GARCH模型 DCC-GARCH模型 溢出指数模型 

分 类 号：F832[经济管理—金融学] O213[理学—概率论与数理统计]