Most of the academic literature on connectedness focuses on stock markets and commodity spot markets. However, there is still much to say about the connectedness among commodity futures markets at different expiration dates, as this part of the literature is as yet small and inconclusive. This paper builds on the existing literature by focusing on connectedness among a set of ten commodity futures markets (including energy, agriculture, and metal sectors) at different maturities, the global equity market and three different sources of uncertainty (financial, economic, and geopolitical) over the period 2000–2022. In doing so, we estimate a combination of complementary connectedness metrics based on the work of Diebold and Yilmaz (2012), which enables measuring average, and the works of Chatziantoniou et al. (2021) and Ando et al. (2022) which enable measuring connectedness under different market circumstances (i.e., low, median and high quantiles). The analysis provides evidence of the variable aspect of connectedness across commodities and uncertainty measures assessed across different quantiles. The average directional connectedness network suggests that commodity futures markets within the same category are significantly sensitive to each-other. However, interdependencies between commodities belonging to different categories are relatively lower. The average total connectedness across quantiles provides evidence of (i) a clear symmetric pattern at the extreme lower and upper quantiles, and (ii) an increase in connectedness with the magnitude of extreme negative and positive shocks. The time-varying analysis indicates that connectedness increases at all quantiles during periods of high market stress, but with relatively higher intensity at the lower quantiles. Additionally, the structure and magnitude of connectedness at the extremes – upper and lower quantiles – differs from the pattern of connectedness at the median quantiles.

大多数有关连通性的学术文献都集中在股票市场和商品现货市场。然而，关于不同到期日的商品期货市场之间的关联性仍然有很多话要说，因为这部分文献还很小而且还没有定论。本文以现有文献为基础，重点关注十种不同期限的商品期货市场（包括能源、农业和金属行业）、全球股票市场以及三种不同的不确定性来源（金融、经济和地缘政治）之间的关联性。 ）2000年至2022年期间。在此过程中，我们根据 Diebold 和 Yilmaz (2012) 的工作（可以测量平均值）和 Chatziantoniou 等人的工作来估计互补连通性指标的组合。（2021）和安藤等人。（2022）能够衡量不同市场环境（即低分位数、中分位数和高分位数）下的连通性。该分析提供了跨商品关联性的可变方面以及跨不同分位数评估的不确定性度量的证据。平均方向连通性网络表明，同一类别的商品期货市场彼此之间显着敏感。然而，属于不同类别的商品之间的相互依赖性相对较低。分位数之间的平均总连通性提供了以下证据：(i) 在极端下分位数和上分位数处存在明显的对称模式，以及 (ii) 连通性随着极端负面和正面冲击的幅度而增加。时变分析表明，在市场压力较大的时期，所有分位数的连通性都会增加，但较低分位数的强度相对较高。此外，极端（上分位数和下分位数）的连通性结构和大小与中分位数的连通性模式不同。