We propose a new measure to establish price leadership among multiple related price series using a multivariate Markov chain model. This new measure, the price leadership share (PLS), can easily be calculated when price series are related but not fully cointegrated (e.g. there is a fractional cointegration and the unit root test fails) and with more than two price series simultaneously, offering advantages over the existing price discovery measures. In addition, we propose a price leadership concentration index to help the comparative analysis. The measure is tested on six gold contracts, including spot, futures, and ETF, with a global coverage over a 2-year period. Results show that gold futures contracts, mainly the US contract (CME futures), have a major role in the price discovery function confirming the previous literature’s findings. Overall, the PLS measure overcomes the limits of other existing price discovery measures.

我们提出了一种新的方法，即使用多元马尔可夫链模型在多个相关价格序列之间建立价格领导地位。当价格序列相关但未完全协整时（例如，部分协整和单位根检验失败），并且同时具有两个以上的价格序列，可以轻松计算出这一新的指标，即价格领导份额（PLS）。超过现有的价格发现措施。此外，我们提出了价格领先集中度指数以帮助进行比较分析。该措施已针对包括现货，期货和ETF在内的六种黄金合约进行了测试，这些合约在两年内覆盖全球。结果表明，黄金期货合约，主要是美国合约（CME期货），在证实先前文献发现的价格发现功能中起主要作用。全面的，