Modelling the data-generating mechanism of China’s commodity market by identifying hidden information flow regimes

Abstract The commodity market reflects extensive macroeconomic information, and quantifying these information flows through fluctuations in commodity price indices can enhance market monitoring and trend forecasting. This paper employs a high-order hidden Markov chain to quantify latent macroeconomic information flows in China’s Commodity Futures Index (CFI) over the period from June 25, 2004, to January 31, 2023. Our empirical findings offer valuable insights for investors and regulators. First, hidden macroeconomic information flows operate in either a high or low volatility regime. During the high-volatility regime, the CFI exhibits larger swings and more frequent jumps, underscoring the role of latent information in short-term demand and supply dynamics. Second, commodity markets exhibit heterogeneous data-generation mechanisms (i.e., the industry, metal, and energy markets are more sensitive to shocks than the agriculture market). Third, estimated macroeconomic information flows serve as explicit leading indicators for key macroeconomic variables.