A1295
Title: Connectedness between climate and commodities: A new measure using mixed-frequency VECM
Authors: Azra May Kabiri - Macquarie University (Australia) [presenting]
Yanlin Shi - Macquarie University (Australia)
Zhuo Jin - Macquarie University (Australia)
Abstract: The linkages between climate variables and commodity prices are examined using a new ratio-based measure of connectedness. Our dataset includes monthly temperature anomalies and El Nino-Southern Oscillation (ENSO) indices with daily energy and agricultural futures prices. To accurately capture the linkages, we develop a ratio-based connectedness measure, which leverages recently modified impulse response functions (MIRFs) and is tailored to mixed-frequency settings. The MIRF refines the forecast-error variance decompositions, which are the critical input to compute connectedness. The ratio-based metric further mitigates potential bias when estimating connectedness in mixed-frequency VAR-type models. Monte Carlo simulations demonstrate that our measure yields significantly lower bias and greater efficiency relative to existing connectedness metrics. We then implement this metric within a generalized VECM that accommodates both non-stationarity, stationarity, and cointegration among variables. We show that climate variables serve as net transmitters of shocks to commodity futures markets, even after controlling for gold price fluctuations. Specifically, both temperature anomalies and ENSO events exert significant influences on commodity prices, with temperature anomalies displaying the most pronounced effects. These findings persist across various sensitivity analyses.
