黄金价格和美元之间联动关系的研究一直非常重要,已有文献大多基于线性相关系数、Granger因果关系检验和Copula模型等进行分析。本文则基于分位点相协测度对上述相依关系进行实证研究,并基于时变分位点相协回归模型,基于局部多项式方法给出黄金价格和美元指数之间的上尾和下尾的时变相依关系。实证结果证明,在大部分时间,黄金价格和美元指数都存在负相关关系。然而,在发生非常事件的特殊时期,二者的下尾和上尾相依系数会明显大于1,即美元与黄金价格呈现正向关系,如政治格局变换、战争爆发和经济大波动等时期,在不同的分位点水平上都得到了相同的结论。另外,本文还结合相应的政治和经济背景对黄金价格和美元指数的相关关系的结构性变化进行了原因分析,这对于投资者把握黄金的走势和风险度量具备一定的参考意义。
The research on the price of gold and the US dollar index mainly choose the linear correlation coefficient, Granger causality test or Copula models. But there are some limitations of them. Different from previous studies, in this paper, time-varying quantile association regression model is constructed. We use the model and the method of local polynomial regression to study the time-varying tail dependence between the price of gold and the US dollar index at different quantile levels. The empirical results show that the price of gold and the dollar index are negatively correlated most of the time. There is a positively related dependence during the extraordinary times at different quantile levels, like the transform of the political landscape, the outbreak of war and the large fluctuations in the economy of the period. Finally, we analyzed the correlation between the gold price and the US dollar index combined with the corresponding political and economic context. These empirical results have great significance for investors to recognize the trend of gold.