This paper investigates the tail behavior patterns of commodity assets, the risk exposure of these assets, and how they rank given their safe haven properties. We use state-of-the-art dynamic generalized autoregressive score models to jointly estimate tail risk measures for ten commodity assets (aluminum, copper, crude oil, gasoline, gold, heating oil, lead, soybeans, tin, and wheat) over the period from September 14, 2011 to June 30, 2021. Our in-sample findings suggest that aluminum outperforms gold as a safe haven in both pre- and COVID-19 times. The out-of-sample results confirm that aluminum retains its leading role during the COVID-19 pandemic. These findings bear implications for constructing well-diversified portfolios which is vital for investors, portfolio managers, and financial advisors, and for policymakers to design policies that ensure financial stability during periods of market turmoil, such as the COVID-19 pandemic.

本文研究了商品资产的尾部行为模式、这些资产的风险敞口，以及它们如何根据其避险属性进行排名。我们使用最先进的动态广义自回归评分模型来联合估计 10 种商品资产（铝、铜、原油、汽油、黄金、取暖油、铅、大豆、锡和小麦）的尾部风险度量。从 2011 年 9 月 14 日到 2021 年 6 月 30 日期间。我们的样本调查结果表明，在 COVID-19 之前和 COVID-19 时期，铝作为避风港的表现都优于黄金。样本外结果证实，铝在 COVID-19 大流行期间保持其主导作用。这些发现对构建对投资者、投资组合经理和财务顾问至关重要的多元化投资组合具有重要意义，