A Pattern New in Every Moment: The Temporal Clustering of Markets for Crude Oil, Refined Fuels, and Other Commodities

Abstract

A suite of clustering methods, applied to the matrix of conditional volatility by trading days and individual assets or asset classes, can identify critical periods in markets for crude oil, refined fuels, and other commodities. Unsupervised machine learning provides a viable alternative to rules-based and subjective definitions of crises in financial markets and the broader economy. Five clustering methods—affinity propagation, mean-shift, spectral, k-means, and hierarchical agglomerative clustering—can identify anomalous periods in commodities trading. These methods identified the financial crisis of 2008–09 and the initial stages of the Covid-19 pandemic. Applied to four energy-related markets—Brent, West Texas intermediate, gasoil, and gasoline—the same methods identified additional periods connected to events such as the September 11 terrorist attacks and the 2003 Persian Gulf war. t-distributed stochastic neighbor embedding facilities the visualization of commodity trading regimes. Future applications of these methods include the definition of bull and bear markets and the identification of recessions and recoveries in the real economy.