Non-linearities in the relationship of agricultural futures prices
采用非线性平滑转换方法，研究七种农产品期货的近月与次近月合约价格关系，发现线性模型忽略重要动态，期货价差在现货溢价时期预测力更强。
The movement of food prices remains a controversial issue owing to the intense rise in volatility that has been observed in recent years. Agricultural futures markets have experienced a similar pattern and simplistic linear models seem to be no longer reliable when analysing their functions. Against this background, this study contributes to the literature by adopting a non-linear smooth transition approach to examine the relationship between prices for first and second nearby futures contracts of seven agricultural commodities. Our main objective is to distinguish between contango and backwardation regimes when analysing the relationship between the futures spread and changes in the first nearby futures price. Our findings reveal that a linear framework neglects important dynamics, as futures prices adjust only under specific circumstances, and that the predictive power of the futures spread is much stronger during backwardation regimes.