This paper investigates the long-term common movement of resource and commodity prices. Beyond its unquestionable policy relevance, detecting such common behaviour is empirically challenging. A novel methodological approach is proposed. It is based on a common latent factor hypothesis. This hypothesis is empirically investigated by specifying a FAVAR-MGARCH model combining the main univariate and multivariate stochastic features of these series. The two latent factors move around a zero-mean short-term level and a non-stationary long-run equilibrium level, respectively. Few heterogeneous and mostly unrelated resources and commodities are considered (crude oil, copper, wheat, beef, aluminium and corn). Using IMF monthly prices over the 1980:1–2019:12 period, a Kalman Filter ML estimation is performed. Results suggest that, besides the time-varying price volatility, the last 15 years have seen a slight rise of the long-term nominal prices corresponding to a stabilization of the respective long-term real prices after a period of regular decline. Policy implications seem relevant and deserve further investigation.

本文研究了资源和商品价格的长期共同运动。除了其毫无疑问的政策相关性之外，检测此类常见行为在经验上也具有挑战性。提出了一种新颖的方法论方法。它基于常见的潜在因素假设。通过指定结合了该系列的主要单变量和多变量随机特征的FAVAR-MGARCH模型，对这一假设进行了经验研究。这两个潜在因素分别在零均值短期水平和非平稳长期均衡水平附近移动。很少考虑种类繁多且大多无关的资源和商品（原油，铜，小麦，牛肉，铝和玉米）。使用IMF在1980：1–2019：12期间的月度价格，进行卡尔曼滤波器最大似然估计。结果表明，除了随时间波动的价格波动外，过去15年中长期名义价格略有上升，这与经过一段时间的定期下降后各个长期实际价格的稳定相对应。政策含义似乎是相关的，值得进一步调查。