Market interdependence among commodity prices based on information transmission on the Internet

• The impacts of information from Internet on commodity prices are analysed.
• The changes of concern in the Internet can granger cause market prices.
• The spillover effects of information transmission on market prices are verified.
• The conditional correlation among commodity prices varies dramatically over time.

Human behaviour on the Internet has become a synchro-projection of real society. In this paper, we introduce the public concern derived from query volumes on the Web to empirically analyse the influence of information on commodity markets (e.g., crude oil, heating oil, corn and gold) using multivariate GARCH models based on dynamic conditional correlations. The analysis found that the changes of public concern on the Internet can well depict the changes of market prices, as the former has significant Granger causality effects on market prices. The findings indicate that the information of external shocks to commodity markets could be transmitted quickly, and commodity markets easily absorb the public concern of the information-sensitive traders. Finally, the conditional correlation among commodity prices varies dramatically over time.