Volatility Spillovers in Commodity Markets: A large t-vector autoregressive approach

Luca Barbaglia, Christophe Croux, Ines Wilms

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Prices of commodities have shown large fluctuations. A high volatility of one commodity today may impact the volatility of another commodity tomorrow. As such, agricultural and energy commodities are closely dependent due to the expansion of the biofuel industry. We study volatility spillovers among a large number of energy, agriculture and biofuel commodities using the vector auto regressive (VAR) model. To account for the possible fat-tailed distribution of the model errors, we propose the t-lasso method for obtaining a large VAR. The t-lasso is shown to have excellent properties, and a forecast analysis shows that the t-lasso attains better forecast accuracy than standard estimators. Our empirical analysis shows the existence of volatility spillovers between energy and biofuel, and between energy and agricultural commodities.
Original languageEnglish
Article number104555
Number of pages11
JournalEnergy Economics
Volume85
Early online date2 Nov 2019
DOIs
Publication statusPublished - Jan 2020

JEL classifications

  • c32 - "Multiple or Simultaneous Equation Models: Time-Series Models; Dynamic Quantile Regressions; Dynamic Treatment Effect Models"
  • q02 - Global Commodity Crises
  • c58 - Financial Econometrics

Keywords

  • Lasso
  • Multivariate t-distribution
  • Vector AutoRegressive model
  • Volatility spillover
  • commodities
  • forecasting
  • REGRESSION
  • Vector autoregressive model
  • Commodities
  • SYSTEMIC RISK
  • Forecasting
  • OIL PRICES
  • RETURN
  • TRANSMISSION
  • REALIZED VOLATILITY
  • VARIANCE

Cite this