关键词: 中国黄金市场, 非线性动力系统, 递归图, 金融风险识别

Abstract: The method of chaos recurrence plot is used to analysis the dynamic properties of Chinese golden market and realizes crisis identification and prediction. The starting point for the study comes from chaotic dynamics analysis with the aim of testing whether the data from Chinese golden market may generate within the framework of a chaotic system. The results confirm that Chinese golden prices time series display long memory and can be modeled by the dynamical systems. In the second part, three different types of crisis phases among the golden market are come from. Then, the recurrence plot is developed to investigate the m-dimensional phase space trajectory through a two-dimensional representation of its recurrences. Finally, the similarity between recurrence plots is measured using RPCD distance to identify the crisis hidden behind the golden market. The empirical study shows that RP with RPCD similarity deduces the false positive effectively compared with Euclidean distance and Dynamic Time Warping.

Key words: Chinese golden market, non-linear dynamics, recurrence plot, financial crisis identification