We present evidence on breaks, stationarity, trends, dynamic correlations, and cycles in prices for eight agricultural and industrial commodities over seven centuries. We utilise a sequential testing procedure that endogenously determines structural changes in the trend and level of each series. We find that breaks in the slope and the trend function can be attributed to major historical events such as the linking of the ‘Old’ and ‘New’ Worlds during the sixteenth century, the early free trade agreements in the nineteenth century and World War I, World War II and the Depression in the twentieth century. Give the importance of crude oil and its impact on agricultural and industrial commodities, we find a time-varying correlation structure which has intensified in the post-World War II period. We also find that three quarters of the commodity prices contain a unit root and that mean periodicity of cycles range between 30 and 35 years, depending on the commodity. These results suggest that during periods of rising or falling commodity prices, prices may not return to the long-run trend for extended periods. Consequently, there is a need for long-term planning initiatives that smooth commodity prices around the trend, with particular attention given during oil price shock.

我们提供了七个世纪以来八种农业和工业商品价格的中断、平稳性、趋势、动态相关性和周期的证据。我们利用顺序测试程序，内生地确定每个系列的趋势和水平的结构变化。我们发现斜率和趋势函数的中断可以归因于重大历史事件，例如 16 世纪“旧”和“新”世界的联系、19 世纪早期的自由贸易协定和第一次世界大战，第二次世界大战和二十世纪的大萧条。考虑到原油的重要性及其对农业和工业商品的影响，我们发现了一种时变相关结构，这种结构在二战后时期有所加强。我们还发现，四分之三的商品价格包含单位根，平均周期周期在 30 到 35 年之间，具体取决于商品。这些结果表明，在商品价格上涨或下跌期间，价格可能不会在很长一段时间内恢复到长期趋势。因此，需要制定长期规划举措，使商品价格顺应趋势平稳，尤其是在油价冲击期间给予特别关注。