AI 中文总结
本研究提出两阶段“分而治之”框架，分解商品风险的微观、市场及宏观来源，构建风险强度指数，发现市场风险占比最高且风险高度集中，为投资者和政策制定者提供风险诊断工具。
AI 中文摘要
我们通过区分微观、市场层面和宏观经济层面的来源来研究商品风险的解剖结构。我们开发了一个两阶段的“分而治之”框架，允许这些风险来源的敏感性在不同商品之间有所变化，同时将宏观经济层面的风险视为潜在变量。第一阶段使用去因子化的工具变量估计来恢复商品对微观和市场条件的特定敏感性。第二阶段将主成分分析与高维变量选择相结合，以识别宏观金融风险的可观测表示。然后，我们构建了风险强度指数（RIIs），该指数将估计的敏感性与当前风险状况相结合，以共同尺度量化每个风险来源的相对重要性。市场风险平均而言是最大的组成部分，约占总风险强度的五分之二，对于能源商品而言则超过一半。风险强度在单个商品中也高度集中：前20%的商品约占微观和市场风险强度的一半，而宏观风险的分布则更为广泛。风险的构成在不同行业和时间段之间差异显著，市场风险在商品市场压力时期尤为突出。微观和市场RIIs还包含有关未来波动率和绝对收益的信息。这些发现为投资者、风险管理者及政策制定者提供了关于商品风险集中何处、哪些风险层级最为重要以及其重要性如何随时间变化的诊断。更广泛地说，我们的“分而治之”框架提供了一种灵活的方法，用于在共同风险为潜在变量的环境中分解分层风险。
英文摘要
We study the anatomy of commodity risk by distinguishing micro, market-level, and economy-wide sources. We develop a two-stage "divide-and-conquer" framework that allows sensitivities to these risk sources to vary across commodities while treating economy-wide risk as latent. The first stage uses defactored instrumental-variable estimation to recover commodity-specific sensitivities to micro and market conditions. The second combines principal components with high-dimensional variable selection to identify an observable representation of macro-financial risk. We then construct Risk Intensity Indices (RIIs), which combine estimated sensitivities with prevailing risk conditions to quantify the relative importance of each risk source on a common scale. Market risk is the largest component on average, accounting for about two fifths of total risk intensity and more than half for energy commodities. Risk intensity is also highly concentrated across individual commodities: the top 20% account for approximately half of micro and market risk intensity, whereas macro risk is more broadly dispersed. The composition of risk varies substantially across sectors and over time, with market risk becoming particularly prominent during episodes of commodity-market stress. Micro and market RIIs also contain information about future volatility and absolute returns. These findings provide investors, risk managers, and policymakers with a diagnostic of where commodity risk is concentrated, which risk layers are most important, and how their importance changes over time. More broadly, our divide-and-conquer framework provides a flexible approach to decomposing layered risk in settings where common risk is latent.
