Copper price determination: fundamentals versus non-fundamentals

TLDR: In this paper, the authors developed an empirical model for the copper market, consisting in a vector autoregression (VAR) with 16 variables, including both fundamentals and non-fundamentals.

Abstract: In mineral economics, there is quite broad consensus that market fundamentals (physical supply and demand) explain commodity price fluctuations, particularly in the medium and long term. However, following the recent price boom, some dissent has arisen about the role played by non-fundamentals such as liquidity or money supply in key countries and regions, the financialization of commodities and, particularly, financial speculation. This paper develops an empirical model for the copper market, consisting in a vector autoregression (VAR) with 16 variables, including both fundamentals and non-fundamentals. Since the variables’ impact probably changes over time, the 20 years studied are divided into three periods (1995–2003, 2003–2008, and 2008–2015) separated by two structural breaks related to the Chinese boom and the financial crisis/Great Moderation. The results show that, although the fundamentals are relevant in all the periods analyzed, liquidity and other macroeconomic variables are also necessary in order to understand the level of copper prices and their fluctuations. In the case of financial speculation, the results indicate that its impact was significant only in 2003–2008 and, even then, was smaller than that of the fundamentals and macroeconomic variables, explaining around 9% of the price increase in this period. The results support the conclusion that, for the purposes of modeling and forecasting, current models based only on the fundamentals cannot fully explain price dynamics which are shown to be, in general, more complex than has been assumed by mainstream mineral economics.