Dynamic Risk Frameworks for Commodity Portfolio Optimization

Abstract

Commodities are an important yet poorly understood asset class, but outsized losses and gains in commodities in recent years have garnered public attention. Partly in reaction to financial market crashes, the suite of risk management tools has expanded considerably since 1996 when J.P. Morgan published Value-at-Risk. In parallel, significant advancements in financial forecasting have been made since Engle’s ARCH model. We compare these and other tools in extreme commodity market environments and observe that dollar-denomination effects, high volatility, and high correlation adversely affect their performance. Our results have implications for investors, commercial hedgers, and regulators tasked with reducing systemic risk.